Opinion
No. H010375
4-25-1995
Previously published at 34 Cal.App.4th 460, 39 Cal.App.4th 343, 43 Cal.App.4th 1440, 48 Cal.App.4th 740 34 Cal.App.4th 460, 39 Cal.App.4th 343, 43 Cal.App.4th 1440, 48 Cal.App.4th 740, 63 USLW 2719 The PEOPLE, Plaintiff and Respondent, v. Gustavo MARLOW, Jr., Defendant and Appellant.
Kyle Gee, Oakland, for appellant. Daniel E. Lungren, Atty. Gen., George Williamson, Chief Asst. Atty. Gen., Ronald A. Bass, Sr. Asst. Atty. Gen., Ann K. Jensen, Supervising Deputy Atty. Gen., Enid A. Camps, Deputy Atty. Gen., for respondent.
The PEOPLE, Plaintiff and Respondent,
v.
Gustavo MARLOW, Jr., Defendant and Appellant.
April 25, 1995.
As Modified on Denial of Rehearing May 24, 1995.
Certified for Partial Publication. *
Review Granted July 20, 1995.
Kyle Gee, Oakland, for appellant.
Daniel E. Lungren, Atty. Gen., George Williamson, Chief Asst. Atty. Gen., Ronald A. Bass, Sr. Asst. Atty. Gen., Ann K. Jensen, Supervising Deputy Atty. Gen., Enid A. Camps, Deputy Atty. Gen., for respondent.
COTTLE, Presiding Justice.
I
INTRODUCTION
In the published portion of this appeal, we hold that DNA typing evidence undertaken pursuant to the restriction fragment length polymorphism (RFLP) process is admissible in a court of law because the processes of matching DNA and of assigning the match a statistical significance have achieved general acceptance in the relevant scientific community. 1 We also hold that the rule recently reaffirmed in People v. Leahy, supra, 8 Cal.4th at p. 595, 34 Cal.Rptr.2d 663, 882 P.2d 321, that "once a trial court has admitted evidence derived from a new technique and the decision is affirmed on appeal in a published opinion, it will become precedent controlling subsequent trials" applies to DNA typing evidence using the RFLP process. II
BACKGROUND
A. The Non-DNA Evidence
1. The Martha D. Homicide
In April 1988, 22-year-old Martha D. lived with her parents and sister in rural San Benito County. She was attending community college and had a job at a Quick Stop store in Hollister. Her 17-year-old sister Yolanda worked at the Wendy's restaurant "next door" to Quick Stop.
Martha and her sister Yolanda drove to work together on Saturday April 2, 1988, the day before Easter. Martha's shift began at 4 p.m. while Yolanda's began an hour later. The sisters' shifts both ended at midnight, when Martha would drive them home. At about 9:15 that evening, the two met and discussed getting some food when they got off work.
During Martha's shift, one of her coworkers noticed a young Hispanic man about 19 or 20 years old come into the market and buy a single rose, which he gave to Martha.
Just after midnight, Martha left work. Usually her coworker, Gary Youtsey, would watch her get into her car. Martha would wave and he would know everything was all right. That night, however, was particularly busy so he did not see Martha leaving. He did notice her driving her car out of the parking lot, but she did not wave to him as usual. Youtsey did not see anyone in the car with Martha.
A few minutes later Yolanda got off work and went looking for Martha. She noticed the car was not in the parking lot. Yolanda went into the Quick Stop where Youtsey told her that Martha had already left. Yolanda was incredulous. Her sister had never before left her stranded without any way home.
Yolanda called her brother at 12:15. He picked her up and drove her to their parents' home. By this time, it was 1 a.m., and Martha was not at home. That was very uncharacteristic of Martha.
Marilyn Klauer lived across the street from the San Benito High School library. At 12:20 a.m., she heard a car pull up and noticed the reflection of its headlights in her window. Klauer heard a female's "loud piercing scream" and a male yelling. The screaming and yelling continued for over a minute. Afterwards, Klauer heard a "sobby cry" followed by a car door slamming. She looked out her bedroom window and saw a light colored car with its lights on with the passenger side door wide open. Klauer did not report the incident to police as she often heard noisy teenagers because of her proximity to the high school.
As Martha's brother Joel was returning home after taking Yolanda to his parents' home, he spotted Martha's car next to the San Benito High School library. Inside the car, which still had its lights on, Joel saw Martha's purse and a rose. He called his father and police.
Police found Martha's nude body between two temporary classrooms on the high school campus, approximately half a block from her car. She appeared to have been strangled with a portion of her blouse, her other clothes having been thrown onto a nearby roof, wrapped in a bundle.
Martha's face was swollen and bruised, and the front of her body was scratched and covered with grass stains. It appeared that she had been dragged face down by her feet from the place she was attacked to the area between the two classrooms where she was found. There was also evidence of vaginal damage, including evidence of forcible entry. A vaginal swab contained non-motile sperm.
Yolanda testified that she had known defendant Gustavo Munoz Marlow, Jr. since seventh grade. (Defendant was one year ahead of her at school.) It was Martha who had introduced them. Martha had been a teacher assistant at defendant's school. Martha and Yolanda had invited defendant to their church and had given him rides on occasions. Martha was planning to participate, with defendant, in a fund-raiser for Hispanic youths at the time she was killed.
One of defendant's friends, Robert Villegas, testified that he had played a game called Ouija board with defendant before Easter 1988. In the game, "spirits" allegedly spell out messages, but Villegas admitted that a player could, if he wanted to, direct the movement of the game piece. One time defendant got a message telling him "to kill and rape a girl [named Jill Alexander] to get more power." After that episode, Villegas told defendant he did not want to play the game any more. Villegas and another friend of defendant's named Eddie King said that defendant told them he was in love with a girl named Martha.
Marilyn Klauer testified that less than three weeks after the San Jose Mercury News described her as a witness in Martha's murder, defendant appeared at her backyard gate and attempted to get in. He stopped when Klauer's dog came out and tried to bite him. Klauer asked defendant if she could help him with something and defendant responded by asking her if she spoke Spanish. Klauer was positive the person who attempted to enter her yard was defendant because of his distinctive "unusual, ... almond-type" eyes.
Hollister police originally suspected an individual named Raul Zamudio of Martha's murder. Zamudio had an abnormal fascination with Martha's case and had told a psychic that he had been talking to Martha "beyond the grave." However, Zamudio was over 40 years old (whereas the male who gave Martha the rose was 19 or 20) and he had had a vasectomy (whereas the male who raped Martha left semen). Subsequent DNA tests ruled out Zamudio as a suspect.
Defendant's fingerprint was found on the outside of the front passenger-side door of Martha's car. However, gold earrings Martha had been wearing on the evening she was murdered were never found.
Over three years later, while defendant was in jail, he passed letters back and forth to a fellow inmate. In one of the letters, defendant admitted he found Martha's car and pulled out a gun and told her to drive.
2. The Lisa K. Homicide
The same Marilyn Klauer who saw Martha's car from her window on early Easter morning, 1988, owned the K & S market in Hollister. Sixteen-year old Lisa K. worked part-time at that market.
On July 26, 1988, Lisa drove to work in her parents' 1975 Chevrolet van, which she parked in a dark corner of the parking lot. About 8:30 p.m., while Lisa was at work, defendant entered the Sunrise Video store, directly across the street from the K & S market. There he talked to his friend Ginger Thompson. When Thompson had earlier told defendant about her marital problems, defendant told her he "could get [her husband] killed." He also told Thompson about the Ouija Board telling him to kill a girl named Jill Alexander to "get powers from it."
Sophia Bueno also saw defendant at the video store talking to Thompson. Later, at 9:15 p.m., as she waited for a ride from her boyfriend, she saw defendant outside at a pay phone.
Robert Kinslow, who had been on the same junior high school wrestling team as defendant, saw defendant in front of the K & S market somewhere between 9 and 9:30 p.m., when he was on an errand for his employer. Fifteen or twenty minutes later, when Kinslow was returning to work, he saw that defendant was still there "walking around in a circle." Kinslow remarked, "[are you] still [ ]here" and defendant acknowledged, "Yeah."
Another schoolmate, Eric Willhide, also saw defendant sitting on an island in the middle of the K & S parking lot at around 9 p.m.
Lisa got off work at about 9:55, then bought a pie and carton of milk for $.54. The grocery receipt for these items, which was found in Lisa's smock pocket, was time-stamped 10 p.m. exactly.
John Klauer, whose family owned the K & S market, saw Lisa in her van in the parking lot as he was leaving. She waved at him, so he drove off. Moments later, he noticed that Lisa was still in the van, that she had not turned her headlights on, and that she had not started moving. He became concerned and turned around to check on her. By the time he got back to the parking lot, however, the van was gone.
Lisa's father, a highway patrolman, and her mother went to look at some furniture on the evening of July 26, 1988. At around 10 p.m., they drove by the K & S parking lot. They were worried that the van's rear door might have been left unlocked, and they decided to check on it. However, when they were about 250 feet away, they saw the van leave the parking lot. They then left to look at the furniture.
When they got home at 11 p.m., Lisa was not there.
At around 11:20 p.m., Joanne Lamberson saw a van drive into her cul-de-sac and shut off its lights. She saw a man get out, close the door, and walk away in the direction of Sunnyslope Road. Defendant's apartment was in that direction, approximately 725 feet from where the van was parked. When Lamberson went to work the next morning, the van was still there.
In the meantime, Lisa's parents were calling all her friends and acquaintances to ascertain her whereabouts. Eventually, they called county communications. Anne Perez was a friend of the family, and she waited with them to hear word of Lisa. After not hearing anything and to relieve her stress, Perez decided to take a bike ride. At 7 a.m. she discovered the van and called Lisa's parents. Lisa's father arrived shortly thereafter and saw blood on the seat and Lisa's clothes in the back. The van was unlocked and the keys were on the seat.
Just a few minutes before Perez discovered the van, an individual named Paul Gonzales found Lisa's nude body about three feet off of Quien Sabe Road. He called police. They found blood around Lisa's mouth, right shoulder and groin, and numerous stab wounds to the front and back of her chest and her neck. Seminal fluid was found in and near her body.
Evidence suggested Lisa had been repeatedly stepped on and kicked. A shoe tread found on her neck matched the tread on a pair of Pony shoes later found in defendant's apartment. 2 Evidence also suggested Lisa had been run over. There was a tire tread mark on her shoulder and neck matching the van's tire tread. Her skull was fractured and her left ear was partially torn off.
The same shoe tread found on Lisa's neck was also found on a sheet of paper found in the van. The paper was a flier advertising a church fashion show. It had been given to Lisa on July 24, 1988, by her friend Maia Nielsen. The fliers were not printed before that day.
After defendant was arrested and given his Miranda (Miranda v. Arizona (1966) 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694) rights, he was asked where he had been on the evening of July 26, 1988. He said he had been with his downstairs neighbors and that he had not been to the Sunrise Video store, the K & S market, or the parking lot in front of the K & S market. He said he did not touch any vehicles or vans in the parking lot.
At trial, a fingerprint expert testified that defendant's palm prints matched two latent prints found in Lisa's van.
In the jail house letters mentioned in connection with the Martha D. homicide, defendant also stated: "I planned it out. I was watching Lisa and saw what she drove. I then made my decision to kill her. I had to kill a cop's relative, so I picked Lisa because her dad was a highway patrolman. So I planned it out, and I hid in the back of her van, and then when she came out and started the van, I jumped out with a gun and scared her, told her where to drive."
3. Conventional Serological Evidence
Department of Justice criminalist Marie Samples performed "conventional" serological tests on the bodily fluid and vaginal swabs taken from Martha and Lisa, on the semen found on Martha's pubic hair, and on blood samples taken from the two suspects, defendant and Raul Zamudio. These tests included (1) "ABO" typing (i.e., determining whether the suspects' and victims' blood types were A, B, AB, or O); (2) phosphoglucomutase (PGM) subtyping (determining which of 10 different subtypes involving combinations of PGM factors 2+, 1+, 1-, and 2- applied); and (3) ascertaining the suspects' and victims' secretor status (i.e., whether he or she secretes his/her ABO factor in saliva, vaginal fluid or sperm, as does 80 percent of the population).
With respect to the Martha D. sexual assault kit, Samples stated that "the only thing you could safely exclude was someone of type B or type AB." Zamudio was type B and therefore could not have been the sperm donor. On the other hand, defendant was type O and his PGM type was consistent with the PGM type found on the semen on Martha's pubic hair. Samples stated that 59.2 percent of the Hispanic population in San Benito, Monterey, and Santa Cruz Counties had the same ABO type as defendant; 69 percent were secretors, as was defendant, and 28 percent had the same PGM type. Multiplying these frequencies together, Samples estimated that 13 percent of the Hispanic population (and 22.7 percent of the general population) in the three counties would have the same ABO and PGM types as defendant.
With respect to the Lisa K. sexual assault kit, Samples could detect no ABO type or PGM type other than the victim's. Lisa was on her menstrual period when she was killed, and her ABO and PGM activity apparently "overwhelm[ed]" that of the donor. B. Procedural History
On August 2, 1988, San Benito County police obtained a warrant to search defendant's residence and person in connection with an unrelated case, the rape of a 14-year-old girl named Amy L. 3 The warrant was executed the same day and defendant's blood and fingerprint samples were taken.
The very next day, police arrested defendant for the attempted rape of another victim, Karen C. 4
On December 13, 1988, defendant was charged in the instant case with the murders, kidnappings, rapes, and robberies of Martha D. and Lisa K.
The preliminary hearing commenced on June 12, 1989, and, with interruptions, continued until November 16, 1989. Much of the time was consumed by a Kelly/ Frye hearing, 5 lasting from August 8 to November 16, to determine whether evidence of deoxyribonucleic acid (DNA) would be admissible at trial. (The testimony elicited at the Kelly/ Frye hearing will be described in greater detail in Background section D., post.) At the conclusion of the evidence, the trial court held defendant to answer. The court noted, however, that with respect to the Martha D. homicide it would not have bound defendant over except for the DNA evidence.
On March 6, 1990, defendant filed his first discovery motion, requesting materials relating to Cellmark Diagnostics, the laboratory that performed the DNA analysis in this case. The district attorney stipulated to provide the requested materials.
On June 8, 1990, defendant filed a motion to suppress evidence, to wit: his blood and fingerprints, obtained by virtue of the August 2, 1988, Amy L. search warrant. He alleged the affidavit in support of the warrant contained omissions and misstatements and was stale. The court denied the motion.
On July 17, 1990, defendant filed another motion to suppress, pointing out that "numerous items (including three pairs of shoes) were seized which were not described in the warrant." He asked that "all shoes seized as a result of this unlawful search ... be suppressed." The court denied this motion as well.
On September 25, 1990, defendant filed a further discovery motion relative to Cellmark's database. That motion was granted.
On November 14, 1990, the district attorney filed a motion asking the court to take judicial notice of the following materials to establish that DNA RFLP testing satisfied the Kelly/ Frye standard for admitting new scientific evidence: (1) the transcripts (1,863 pages) and court findings from the August-November 1989 Kelly/ Frye hearing held during the preliminary hearing in this case; (2) the transcripts (1,498 pages) and court findings from the June-August 1989 Kelly/ Frye hearing in People v. Barney, Alameda County Sup.Ct. No. H10291; and (3) the transcripts (2,318 pages) and court findings from the March-August 1989 Kelly/ Frye hearing in People v. Axell, Ventura County Sup.Ct. No. CR23911.
At the hearing on the motion, held November 20, 1990, the trial judge announced his intent to read the testimony from the preliminary hearing in this case, which counsel pointed out would take approximately 113 hours to do. The judge stated this would allow him, "in consultation with counsel," to "narrow[ ] ... the number of witnesses or the breadth of testimony we would need" in any future Kelly/ Frye hearing. The court ruled, however, that it would not read the preliminary hearing transcripts in Axell and Barney.
Several discovery compliance hearings took place in the early months of 1991, and in March 1991, defendant filed yet another discovery motion, seeking further materials from Cellmark.
In May 1991, the defense filed a motion to re-inspect Cellmark's Maryland laboratory. The court allowed Drs. Eldred Ribeiro and Paul Hagerman to testify in support of this motion regarding alleged deficiencies in Cellmark's RFLP DNA testing. Compliance hearings continued through the rest of 1991.
On January 13, 1992, the district attorney filed a "Motion for Admission of DNA Evidence" on the ground that Cellmark's DNA tests had passed Kelly/ Frye scrutiny in the recently filed People v. Axell (1991) 235 Cal.App.3d 836, 1 Cal.Rptr.2d 411 opinion, which "is now the law of the State of California and is controlling in this case."
Defense counsel requested an opportunity to present evidence that there had been a change of consensus within the scientific community since Axell. After a hearing on February 18, 1992, the court granted defendant's request but indicated the evidence would have to be presented on declarations, without live testimony. That same day the court granted defendant's motion for change of venue based on the extensive publicity the case had generated in rural San Benito County. It also denied a prosecution motion to rescind the appointment of a second counsel for defendant.
Thereafter, on March 31, 1992, defendant filed his response to the People's motion for admission of DNA evidence, to which he appended over 500 pages of exhibits. Defendant argued that the declarations and exhibits established that "Cellmark's DNA test is no longer accepted, if it ever was" and that he was "entitled to present his evidence at a pretrial Kelly/ frye hearing." the district attorney filed a reply.
On April 13, 1992, the district attorney filed a motion to use four uncharged crimes to establish defendant's identity. 6 The court denied the motion.
On April 14, 1992, the court issued its ruling that DNA evidence would be admissible without further Kelly/ Frye hearings. The court stated: "I know the point is that things have happened since [the Axell court] made its ruling, that things have happened in the scientific community, but whatever has happened has not convinced this Court that it has risen to the level where that opinion, which also deals with the same laboratory that's involved in this case, should be ignored.... And the scientific community generally is always growing and changing and learning new things."
At the beginning of trial, which took place in Tulare County commencing on April 28, 1992, defendant moved to strike the special circumstances allegations pleaded in connection with the homicides on the grounds he was a juvenile when the offenses occurred and Proposition 115 had not yet gone into effect. The court granted his motion.
The trial lasted until late June 1992. Much of the testimony concerned DNA and will be discussed in greater detail in Background section E., post. On June 26, 1992, the jury found defendant guilty of two counts of first degree murder (Pen.Code, § 187), two counts of kidnapping (Pen.Code, § 207, subd. (a)), one count of kidnapping for robbery with respect to Martha (Pen.Code, § 209), two counts of rape (Pen.Code, § 261, subd. (a)(2)), two counts of robbery (Pen.Code, § 211), and one count of vehicle theft (Veh.Code, § 10851). A mistrial was declared on a second kidnapping for robbery count with respect to Lisa.
On July 22, 1992, the trial court ordered CYA to evaluate defendant's amenability to treatment and training at that facility. CYA's response, issued October 5, 1992, was that he was not amenable to treatment and training by their department.
On October 6, 1992, defendant filed a motion for new trial based on the decision in People v. Barney (1992) 8 Cal.App.4th 798, 10 Cal.Rptr.2d 731, and the court's failure to hold a Kelly/ Frye hearing before admitting DNA evidence at trial. The next day he asked that sentencing be postponed so that he could file a Marsden 7 motion. After questioning defendant's two trial counsel, the court denied both the motion for new trial and the Marsden motion.
Defendant was sentenced to "66 years, eight months, plus seven years for a mandatory minimum" in state prison. C. A Brief Description of DNA and the RFLP Process that is Used to Generate a DNA Profile
Every human cell with a nucleus contains 23 pairs of chromosomes, 8 which in turn contain thousands of genes, occupying specific locations on the chromosome. The molecular component of genes is deoxyribonucleic acid (DNA). DNA resembles two parallel, spiral staircases (a double helix), the steps of which are made up of four chemical components--guanine (G), adenine (A), thymine (T), and cytosine (C). Two of these components are present on each step, and they are able to combine with the other two components in a limited number of ways; namely, A will pair only with T, T only with A, C only with G, and G only with C. The combined chemical components are called base pairs, and there are about 3 billion of them in a DNA molecule. The sequence of the base pairs is the same in every cell of a person's body. 9 Furthermore, except in the case of identical twins, the order in which the 3 billion base pairs are arranged is unique to each person.
Of the 3 billion base pairs, most (99.9 percent) are identical among all human beings. (State v. Bible (1993) 175 Ariz. 549, 576, 858 P.2d 1152, 1179.) It is this identity that makes humans look like humans, "rather than dogs or trees." (People v. Castro (N.Y.Sup.1989) 144 Misc.2d 956, 545 N.Y.S.2d 985, 988.) The remaining .1 percent of a person's base pairs (approximately 3 million), however, vary greatly from person to person. These locations of genetic variation are called "polymorphisms."
DNA typing evidence is designed to detect a few highly polymorphic regions (loci) and to distinguish between the varying forms (alleles) of genes or base pair sequences found there. (Spencer v. Com. (1989) 238 Va. 275, 384 S.E.2d 775, 781-783.) The loci scientists have found most useful for DNA testing are the noncoding parts, also known as "junk DNA," which are not known to perform any particular function in the cell. 10 (Caldwell v. State (1990) 260 Ga. 278, 393 S.E.2d 436, 439.) At these particular sites, base pair sequences repeat a varying number of times 11 and hence are called "VNTR," or variable number of tandem repeat sequences.
Restriction Fragment Length Polymorphism analysis (RFLP) is the method used by Cellmark, and several other laboratories, to isolate and analyze polymorphic regions of the human genome 12 that contain VNTRs. It is broken down into six steps: "(1) extraction and purification of the DNA; (2) fragmentation by restriction enzymes; (3) gel electrophoresis in which a positive electrical charge to the bottom of an agarose gel on which a DNA sample is placed causes the DNA to move through the gel from the negative to the positive charge; (4) Southern blotting [ 13] in which the gel and DNA in it are transferred to a nylon membrane for easier handling; (5) hybridization in which the DNA pattern unique to the individual is identified by use of radioactively tagged probes, 'unzipped' DNA segments of a known length and sequence, designed to seek out a predetermined locus in a polymorphic region of the DNA and band with a like segment of DNA; and (6) autoradiography in which a film is developed on top of the nylon membrane, revealing the location of the DNA by bands on the X-ray film, called an autoradiogram or autorad. Use of a single probe produces two bands on the autorad. Thus, running four different probes at the same time results in eight bands." (People v. Axell, supra, 235 Cal.App.3d at p. 846, 1 Cal.Rptr.2d 411; see also OTA report, supra, at pp. 14, 141-153.)
After creating a DNA profile using the RFLP process, the next step is to determine whether the suspect's DNA profile matches the DNA profile of the blood or semen samples taken from the crime scene. If it matches, the final step is to determine the likelihood that someone other than the defendant might have the same DNA pattern. (People v. Barney, supra, 8 Cal.App.4th at p. 806, 10 Cal.Rptr.2d 731; State v. Bible, supra, 175 Ariz. at 579, 858 P.2d at p. 1180.) These last two steps will be discussed in considerably greater detail in connection with defendant's arguments that Cellmark's "match" procedures and Cellmark's "statistical analysis" have not achieved general acceptance in the relevant scientific community. D. August-November 1989 Kelly/Frye Hearing
The first prosecution witness was Stanford University human population geneticist Dr. Anne Bowcock, 14 who uses DNA RFLP analysis for her research in human genetic diseases and population studies. In looking for genetically inherited diseases, she employs "the same approach" used in forensic DNA testing "where you compare alleles ... to look for allele matches." In fact, Stanford's protocol is "very similar" to the one used by Cellmark.
RFLP testing has been used in paternity and immigration testing, prenatal analysis and population studies. Generally, labs using DNA for research or diagnostic purposes have a greater amount of sample than labs using DNA for forensics, but not always. For example, in prenatal tests the samples are very small. "DNA is DNA," Dr. Bowcock explained, "and once you get your result, it doesn't matter how much you had to start with."
Environmental factors such as light, heat, humidity, bacteria and dirt may "contaminate" or "degrade" DNA, in which case no DNA pattern will emerge. Cellmark uses certain controls to alert the tester if something goes wrong. These include "molecular weight size markers" 15 and staining the gel with ethidium bromide before "Southern blotting" to see if the DNA sample is degraded.
Dr. Bruce Kovacs 16 testified that RFLP analysis is "very much" accepted in the scientific community and is widely used. He personally uses DNA RFLP testing for molecular genetics analysis, prenatal genetic analysis, and cancer research. In fact, he stated, "[i]t is hard to think of an area of biology where it is not being applied ... from biology to reproductive medicine, to anthropology," to diseases where genes that cause the disorder are identified by the RFLP process.
His lab's protocol is similar to Cellmark's, except that Cellmark's quality control is even more stringent, probably because Cellmark often has a smaller amount of sample available. Researchers use Cellmark's probes in a variety of situations, including in replication studies and in the analysis of tumors.
Like Dr. Bowcock, Dr. Kovacs testified that if a sample is degraded (for example, by bacteria or other contaminants), it will simply yield no result.
Dr. Robin Cotton 17 testified that the use of DNA probes is well established in the scientific community. She herself has been using them since 1981. DNA probes are used for research and diagnostics (such as mapping the location of genes that cause diseases and identifying whether a parent is a carrier of a particular genetic problem) as well as for forensics.
Cellmark received its probes from Dr. Alec Jeffreys, the scientist who coined the phrase "DNA fingerprinting." Cellmark, in turn, makes the probes available to others for "forensic or other types of testing." In fact, the FBI uses one of Cellmark's probes in its testing. Cellmark studied the probes to ascertain the ability to get DNA from hair, blood, and semen samples, under a variety of adverse environmental conditions. It found that degradation does not change a DNA pattern, but if the degradation is too extreme, it will prevent any pattern from developing.
Cellmark's protocol "has been handed out to a lot of different people," including the Technical Working Group on DNA Analysis Methods (TWGDAM). TWGDAM has published guidelines in the Department of Justice, Federal Bureau of Investigation, Crime Laboratory Digest with which Cellmark is "generally in compliance."
In Cellmark's first proficiency test carried out by the California Association of Crime Lab Directors (CACLD), Cellmark was given 49 samples and was asked to determine which of the samples matched each other. A laboratory technician apparently recombined one sample with a sample from a different source, leading to a false positive. As a result of this error, Cellmark purchased a centrifuge large enough so that samples would not have to be split and recombined.
Dr. Cotton and two other Cellmark employees witnessed technician Tony McNeil's work on the DNA samples in this case, and they reviewed his measurements. Dr. Cotton stated that McNeil followed Cellmark's procedures. The samples tested showed that defendant's DNA banding pattern matched the patterned found in the vaginal swabs in the Lisa K. and Martha D. cases.
Finally, Dr. Cotton opined that Cellmark's method of calculating population frequencies is generally accepted in the scientific community.
Dr. Lisa Forman, 18 who is employed by Cellmark as a population geneticist, testified as to how Cellmark declares a "match" and how it assigns the match a statistical probability.
To determine whether samples "match," Cellmark measures and compares the migration and size of the sample bands with bands of a known size (molecular rulers), which are commercially available molecular weight markers. To "say whether electrophoretically, two fragments are indistinguishable from one another," Cellmark employs the "resolution interval" method "based on published calculations ... by Elder and Southern." (Southern is the scientist for whom Southern blotting is named.) Cellmark determines that fragments match if they are within two resolution intervals.
To determine the match's statistical significance, Cellmark needs to estimate how frequently such a match might arise by chance in the relevant population. To do this, Cellmark compares the size of the bands in the sample DNA pattern with the size of bands found in a previously studied group of individuals (the "data base").
Cellmark's data base included a "[western] Hispanic data base from California comprised of individuals mostly from the Los Angeles Red Cross, but also from the Erwin Memorial Blood Bank in San Francisco[,] an Oriental data base comprised of Asian individuals ... from the West Coast[,] a Caucasian data base comprised predominantly of individuals [from] the Red Cross in greater Delaware[,] a Black data base that is comprised predominantly of individuals from the Detroit Red Cross" and a separate eastern Hispanic data base that includes individuals of Puerto Rican or Caribbean Islands descent. Many prominent academics, including Dr. David Goldman at the National Institutes of Health, Dr. Eric Lander, and Dr. Kenneth Kidd, have reviewed Cellmark's data base.
Cellmark first calculates the frequency with which each band is found in Cellmark's data base and then, by a simple "binomial expansion," calculates the likelihood that an individual would inherit all of the bands found in the DNA sample.
Dr. Forman examined the autoradiograms in this case and compared defendant's banding pattern with the various Cellmark data bases. Depending on which data base was selected, the expected frequency of finding someone with the same DNA pattern as defendant ranged from a high of 1 in 33 million to a low of 1 in 7.4 billion. If one used a more conservative three resolution interval criterion 19 (which diminishes the uniqueness of a band frequency) as compared with the two resolution interval used to determine whether the defendant's DNA banding pattern matches that found at the crime scene, the frequency of defendant's pattern in the Hispanic population is increased from 1 in 500 million to 1 in 33 million. Employing the three resolution interval criteria yields "extremely conservative frequencies" that comport with the consensus opinion of the scientific community that "bin sizes should reflect a more generous [approach to the defendant]."
Dr. Mary-Claire King 20 testified that she has performed RFLP analysis thousands of times in such areas as parentage testing, disease mapping (e.g., breast cancer and systemic lupus) and evolutionary studies in human variation. The RFLP procedure is "the backbone of our research," she explained. She stated that molecular biologists, human geneticists and forensic scientists would probably all agree that DNA probes are appropriate for use in identifying individuals "because there is so much variation between people" at the "hypervariable regions" recognized by DNA probes such as the ones used by Cellmark.
Dr. King stated that if she had a known and unknown biological sample, she would perform RFLP analysis to determine whether they came from the same individual. She said that the only difference between RFLP analysis in parentage testing and RFLP analysis in forensics is in the first step, extraction of the DNA. In forensics, the sample could come from blood, semen, fibroplasts, or hair. However, once the DNA has been extracted, "DNA is DNA is DNA." One of her colleagues at Berkeley, Dr. George Sensabaugh, is a leading expert in extracting DNA from crime scene evidence. The literature on the subject is extensive and extremely well established.
Dr. King has reviewed Cellmark's protocol and was, in fact, one of the scientists Cellmark consulted in preparing its protocol. She said Cellmark has an excellent, indeed the "best[,] reputation" among human population geneticists for the quality of its work.
Dr. King testified that "the nice thing about the RFLP approach is that the safeguards are essentially inherent in the system [--][y]ou simply don't get a result if you haven't done it right.... You can't get a false positive."
She is familiar with the probes Cellmark uses, as she uses the same probes herself. She also is familiar with the method Cellmark uses to declare a "match." She uses similar criteria in her laboratory though Cellmark's procedures are "more sophisticated."
As far as Cellmark's method for assigning a statistical significance to the match, Dr. King testified that Cellmark's data bases are quite reliable. She said Cellmark has been conscientious in obtaining samples "from well defined geographic locales and using samples from ... the same locales as the population to which they need to generalize their results." "Hispanics," like defendant, from California "have [a] different geographic origin, and subtle, though probably real, genetic differences" from "Hispanics" from Miami and from "Hispanics" in New York City.
In 1989 she published an article in The American Journal of Human Genetics expressing her concerns about applying DNA tests to "subgroups" of closely related people, such as the Amish, Pygmies, and people from physically isolated villages who breed only within their group. However, she stated that she knew of no communities in California "that would come close to fitting [her] concerns about a small inbred population."
The final prosecution witness was Cellmark technician Tony McNeil, who explained how he performed the RFLP analysis in this case.
After the prosecution expert witnesses testified, the Kelly/ Frye hearing was continued by stipulation of the parties so that the prosecution experts' testimony could be transcribed and reviewed by the defense experts. Two months later, in November 1989, the Kelly/ Frye hearing recommenced, the defense experts testified and the defense was allowed to cross-examine the prosecution's experts.
The first defense expert called was Dr. Diane Juricek Lavett. 21 Although she had never done "RFLP analysis, per se," she had done "every component step" of the analysis. She admitted she has never been to Cellmark's lab, or to Lifecodes' or the FBI's laboratory (both also perform RFLP analysis) or to any seminars concerning the forensic uses of RFLP. She did, however, have an FBI tape on RFLP.
Dr. Lavett opined it was "premature" to apply DNA profiling to crime scene samples because not enough validation studies and contamination studies had been done and because "statistical evidence is based on very, very scanty population data." She stated that the various labs that do RFLP analysis "support each other" because their financial viability depends on the process being accepted. She also questioned Cellmark's impartiality, noting it was under pressure from the prosecution to make a match and, like all businesses, "Cellmark knows it has to please its customers at some level...."
She was critical of Cellmark's method for declaring a match, stating it does "not define resolution, and yet it's made the basis of a match." She was also critical of Cellmark's CACLD proficiency test results.
With respect to this particular case, Dr. Lavett was particularly critical of Tony McNeil's performance. She said his measurements were so far off that he did not get the same measurement when he measured the band a second time. He also failed to initial the crossouts, as called for in the Cellmark protocol.
Dr. Lavett also felt Cellmark's statistical analysis was flawed because of the issue of subgrouping, specifically "there's a lot of inbreeding ... in the Hispanic population in this town, which does breed within itself, as I understand things, and does not breed with a larger community, nor is there much intermarriage, let's say between the population of this town and the population in Los Angeles...." "[D]ifferent geographic groups," Dr. Lavett explained, "differ in allelic frequency."
Dr. Simon Ford, 22 who derives 80 percent of his yearly income from testifying or consulting for the defense in DNA cases, does not currently work in a laboratory that performs RFLP analysis. However, in the past he worked in the University of California social ecology department in the area of environmental molecular biology. There he looked at, among other things, the DNA content of sewage. He completed the entire RFLP process on one sample, using a different probe than that used by Cellmark. In addition, he performed several of the separate steps involved in RFLP, including extraction, restriction, electrophoresis, Southern blotting, and production of autoradiograms.
Dr. Ford worried that contamination could lead to extra bands showing up on the autoradiograms, which could potentially result in "extra matches." He was also critical of Cellmark's protocol, which he felt was "somewhat cumbersome" and "a bit deficient on the control front." He felt Cellmark had not yet satisfactorily demonstrated the reliability of its work. Further, Cellmark's performance on its CACLD proficiency test demonstrated unequivocally that it was "possible to get a false positive." Additionally, Cellmark appeared to have difficulty discerning when a sample contained DNA from more than one person. It was Dr. Ford's belief that Cellmark did not meet the TWGDAM guidelines. He also expressed concern that there was no licensing or certification process for DNA labs. Finally, he noted that there were serious concerns in the scientific community about subgroups within populations, which undermines Cellmark's statistical analysis of the significance of the match.
Dr. Ford's associate, Dr. William Thompson, 23 was the next defense expert. While he has had no formal training in molecular biology, he has learned a great deal about the subject from Dr. Ford and from Dr. Laurence Mueller, who also teaches at University of California, Irvine. He identified himself as an expert in surveys and social psychology. He noted that a number of scientists in molecular biology, population genetics, and biostatistics had "expressed some concern about the molecular biology of the [RFLP] procedure, or about the interpretation of the results." He named several, including all the defense experts in this case.
He also reported about the results of two surveys he conducted to determine whether molecular biologists felt so certain about the reliability of Cellmark's and Lifecode's DNA analysis that they would "bet their [lives]" on it. Although the respondents had four different answers they could choose among, every one of them selected the two choices stating RFLP analysis was reliable--some indicated it was definitely reliable and they would to "bet their li[ves]" on it, others that it was probably reliable. The magistrate intensely questioned Dr. Thompson on how he could consider these results, where not one respondent thought the analysis was not reliable, as damning to Cellmark. 24 To Dr. Thompson, the answer was that several said they would not bet their lives on the result.
The next defense expert was Dr. Laurence Mueller. 25 He testified that Cellmark's statistical analysis is not generally accepted in the scientific community. Cellmark's techniques have never been published in the peer review literature.
With respect to Cellmark's Hispanic data base, Mueller observed that "there is a consistent excess of [ ] single-banded, or homozygous genotypes." 26 He indicated that one cause of this aberration might be the way Cellmark establishes its "arbitrary bin boundar[ies]."
Another cause "revolves around the existence of genetically differentiated subpopulations within a racial category, such as the Hispanic race." He pointed out that "[p]robably a large portion of [those in Cellmark's San Francisco area Hispanic data base] were Mexican in heritage," but that in addition "we certainly have lots of Central Americans now in California."
Dr. Mueller believed that the existence of population subgroups affects "the ability to use the product rule and multiply genotype frequencies at each of four gene loci, which is what Cellmark does, because ... substructuring can lead to statistical association between genes." The "only way" to deal with substructure, Dr. Mueller opined, "is to collect data where you directly identify subgroups that an individual belongs to, like, is he a Cuban, is he a Mexican, or Puerto Rican, things like that, and keep those data bases separately."
Using a calculation developed by a population geneticist named Nei, Dr. Mueller estimated that there is nearly as much genetic similarity between humans and chimpanzees as between members of the FBI's two Hispanic data bases at one locus. E. DNA Trial Testimony
At trial, the prosecution recalled three of the experts it had brought in to testify at the Kelly/ Frye hearing: Cellmark's research and development manager Dr. Robin Cotton, Cellmark technician Tony McNeil, 27 and biogeneticist Dr. Mary-Claire King. In addition, it called two new witnesses: Dr. Patrick Michael Conneally 28 and Dr. Martin Tracey, Jr. 29
McNeil testified that he was not involved in the CACLD proficiency test that was discussed at the Kelly/ Frye hearing that resulted in the error.
Dr. Cotton testified about the various uses of RFLP testing in research and forensics, the CACLD proficiency tests, and the specific results in the instant case. She acknowledged that after she testified at defendant's Kelly/ Frye hearing, Cellmark was given a second CACLD proficiency test in which it again made an error, this time in the handling. As a result of the error, Cellmark changed its procedures as it had done when the first error was detected in the first proficiency test. Dr. Cotton did not believe that a laboratory's error rate should be factored into the probability of a random match based on population statistics. Rather, she believed those numbers should be presented independently.
As far as the test results in the instant case, the samples obtained from defendant, from Lisa K., and from Martha D. were all run on different days, so there was no possibility that the samples had been mixed, which was the problem in the first proficiency test. Martha D.'s vaginal swabs were run May 26, 1988, and were compared with samples from Raul Zamudio, who had been arrested for Martha's murder. Neither Dr. Cotton nor any other prosecution witness who examined the autoradiograms believed that the pattern found in Martha's vaginal swab matched Zamudio's DNA pattern. Two-and-a-half months later, Cellmark ran the DNA test on defendant's blood sample. The sample yielded eight bands, and matched the pattern found in the Martha D. vaginal swab. Twenty days later, on August 30, 1988, Cellmark ran DNA tests on the Lisa K. vaginal swabs. Again, defendant's pattern matched. Technician Tony McNeil did the initial measurements in the case, and then they were done "again by someone else" to verify the accuracy of the measurements.
Dr. Cotton also explained, as follows, how Cellmark computes the statistical likelihood of a match at each of the genetic markers studied: " Suppose I went out into Visalia, and I took a sample of a hundred people, and I said how many of those people had red hair, and, you know, I got some number; Ten percent of the people had red hair. And then I went out and I said how many people have brown eyes, and I got a number of ten percent. I'm sure more people than that have brown eyes, but anyway. [p] And then if I wanted to know how many people are likely to have both red hair and brown eyes, it would be the product, or ten percent of that original ten percent, or one percent."
Dr. Patrick Conneally testified about Cellmark's DNA RFLP testing and explained the use of population statistics to interpret DNA matches. He has published over 300 scientific articles on population genetics and molecular biology. He has a great deal of experience with population substructures, having worked "extensively" with two of the three major ethnic subpopulation groups in the United States (i.e., the Amish, the Huterites, and the Mennonites.)
Dr. Conneally stated that Cellmark's method for finding the frequency of a band in its population data bases is "conservative" (meaning it favors a defendant) in that it uses "bins" in establish how frequently a pattern appears in the data base. He explained binning to the jury by giving the analogy set forth ante in footnote 19.
He explained that to estimate how common a DNA RFLP profile is, scientists use the "product rule," multiplying the frequencies with which each band is found. He said that this approach is reliable so long as two conditions are met: first, the probes must be genetically independent of each other, and second, there must not be "substantial" population substructuring.
As far as the independence of Cellmark's probes, Dr. Conneally stated it was the opinion of persons at the Human Genome Association, of which he was president, that Cellmark's probes were "absolutely" "genetically independent." He analogized the independent probes to ABO blood groups which are genetically independent from Rh blood groups.
As far as the effect of population substructure on probability estimates, he stated that it was not an issue where there is "some out-marriage" between groups, meaning, for example, that "some Italians marry Germans ... or whatever...." (Emphasis added.) He was very familiar with the December 20, 1991, paper by Lewontin and Hartl (entitled Population Genetics in Forensic DNA Typing, Science, at p. 1745) in which they criticized the product rule for its failure to take into account the possibility of major substructuring in the United States population. Dr. Conneally described the authors as "fruit fl[y]" geneticists and noted that the article appeared side by side with the scholarly article by "very well-recognized" human population geneticists Chakraborty and Kidd. The latter authors found no problem with the use of the product rule.
Dr. Conneally explained that even if there were evidence of complete endogamy (for example, that Italians marry only Italians and never Germans, and that Germans marry only Germans and never Italians), it would not make any difference unless it were shown there were differences in band frequencies between the different groups. However, there "isn't that great a difference between Blacks and Hispanics and Caucasians, [let] alone to have differences within Caucasians." To argue about the calculations is "making mountains out of molehills," and any nonindependence there might be in the data base is more than made up by Cellmark's conservative binning process.
He distinguished a truly substructured society, e.g., the small Karitaina Indian tribe isolated in the Amazon, where the product rule might not be appropriate, with the two Hispanic populations in the United States (the western data base comprised primarily of those of Mexican descent and the eastern data base from the Caribbean and Cuba).
Dr. Conneally pointed out that the National Research Council (NRC) of the National Academy of Sciences had suggested that a hundred samples from each ethnicity be taken so that frequencies could be compared "to see if this [substructuring] is really a problem." Until these studies were done, the NRC recommended that labs be "super conservative" and (1) use the most common of the banding patterns from each of the major races in computing population statistics and (2) apply a conservative "ceiling" adjustment if the frequency were less than 10 percent. 30 The "ceiling principle," Dr. Conneally explained, gives the defendant the benefit of the doubt because in most cases the frequency of a band is considerably rarer than ten percent.
Using a "conservative" method, Dr. Conneally estimated he would expect to find defendant's banding pattern in one in 3.5 million Hispanics, one in 13.7 million Blacks, and one in 7.6 million Caucasians. Using the "more conservative" method that looks for the highest frequency among any race, the likelihood of finding a match was one in 340,000 males. Finally, using the NRC "absolutely ultra-conservative" method, the estimated probability was one in 105,000 males.
Dr. Conneally looked at the autorads in this case and showed the jury how the bands did not match those of Raul Zamudio. He opined it was "impossible" for Zamudio to have been the donor of the sperm found on the Martha D. vaginal swab. In response to a defense argument that tilting the gels might have made a difference in the banding pattern, Dr. Conneally responded, "Well, you could tilt it this way and that way and every way in the world and keep tilting it for the whole number of hours that this was [run], and it wouldn't make any difference."
Dr. Mary-Claire King testified much as she did at the Kelly/ Frye hearing. However, she added that she had been selected as one of the contributors to the National Research Council report on the use of DNA in forensics.
She reviewed Cellmark's work in this case and concluded it was very well done. Like Dr. Conneally, she examined the autorads and determined it was impossible for Raul Zamudio to have been the sperm donor. She testified about the importance of quality control and stated that the authors of the NRC report had decided not to recommend merging the probability estimate of a match with the lab's error rate.
Dr. King also explained the "product rule." She "was actually the person responsible for writing" the section in the NRC report dealing with the variation of the product rule known as the "ceiling principle." She has "gotten a lot of response to this from [her] colleagues in population genetics who feel [she] was much too conservative...." She stated that studies have been done showing that DNA markers are "statistically independent of each other," which is the concern with substructuring. However, she stated there was still "some room for argument on theoretical grounds that we don't have independent of cross markers because ... we don't mate at random in the world." (Emphasis added.)
It was based on these theoretical concerns that the ceiling principle was proposed. She described the principle as "a really unbelievably conservative approach, very, very pro-defense approach." It yields a "vastly too conservative estimate, too high an estimate of the chance that a match would be coincidental, but it's so much too high that ... the real estimate is bound to be better than that."
Using the "ceiling" approach and adding a "95 percent confidence limit" to make it even more conservative, Dr. King estimated that the chance of a coincidental match in the case of Martha D. would be one in 31,000 and in the case of Lisa K. one in 27,000. Dr. King testified that she would rather be extra conservative so that no one would be wrongly accused based on DNA evidence. She stated that in the future when additional information about population subgroups is known, scientists will not have to be so conservative.
Dr. Martin Tracey testified that Cellmark's true error rate on the proficiency tests was less than one percent because approximately 2,500 tests were involved. The "important thing" about the error was that it led to changes in the procedure and to the purchase of a new machine so that mistakes in the future could be avoided. Dr. Tracey believed the "ceiling principle" was designed to address the concerns of early critics, such as Lewontin and Hartl, and that it essentially "overcompensates." He himself had worked primarily with fruit flies and lobsters until the last two years when he began working on human population genetics.
The defense called as witnesses Dr. Lavett and Dr. Mueller, who testified at the Kelly/ Frye hearing, Dr. Eldred Ribeiro, who was earlier permitted to testify at a discovery motion, and Drs. Martin Shapiro, Lyndon Larcom and Aimee Bakken.
Dr. Lavett testified much as she did at the Kelly/ Frye hearing. She admitted she is not a forensic scientist, she has never personally followed Cellmark's protocol, she has never used Cellmark's size markers or probes, she has never attended a forensic science conference or meeting, and she has never published anything concerning forensic DNA analysis. Although she had heard of the renowned Dr. Conneally, she had never read anything by him. Since 1985 she had published articles on "therapy concerning non-abusive adults who were physically abused as children."
Dr. Lavett described the history of RFLP analysis. She said there is a "four stage" process of dissemination of a new technique in the scientific community. The first step is when one or two labs begin working on a technique; they run into problems, find ways to solve them and then prepare a paper, which is peer reviewed, describing the results. In stage two, other labs begin to apply the technique. They too encounter problems, discover solutions, and submit their work for peer review. Stage three involves a wide scale acceptance of the technique, further solving of problems and further peer review. In the final stage there is world-wide acceptance of the technique.
Dr. Lavett was "appalled" by the work being done by Cellmark. She considered DNA forensic techniques to be in "stage one" because only a few labs (Cellmark, Lifecodes, and the FBI) were doing RFLP analysis in 1989. She criticized Cellmark for its lack of duplicate testing, lack of controls, use of ethidium bromide, failure to accurately measure bands, and its performance on the CACLD proficiency tests.
She was particularly critical of Cellmark's "resolution units" criteria for calling a match between samples. On the other hand, she acknowledged that Cellmark's "resolution units" criteria was based on an equation developed by scientists Elder and Southern. "All the scientific community do rely on Elder and Southern at this point in time," she complained, "and so Cellmark is doing nothing other than the rest of the scientific community." She nevertheless felt Cellmark's resolution unit methodology was inappropriate in this case, in part because it had not been validated by any published studies. She noted that Cellmark alone used the "resolution unit" methodology; the FBI used a separate percentage of the band size and Lifecodes used a standard deviation.
She further noted that Cellmark had changed the way it calculated "resolution units" over the years. Prior to 1989, Cellmark would go down one resolution unit from the sample and up one resolution unit from the suspect. If there was an overlap, Cellmark declared a match. Between different gels, Cellmark would go up or down two resolution units. By 1992, Cellmark had changed the way it declared a match. If the gel had fewer than three single locus probe results, a match was declared if the suspect's profile and the crime scene profile were within two resolution units; if the gel had three or four probe results, a match was declared if they were within three resolution units.
By Dr. Lavett's calculations, Cellmark had a 32 percent error rate on its first CACLD proficiency test. She opined that the reason CACLD concluded that Cellmark's error rate was 2 percent was because the examiner met with someone from Cellmark and that made the conclusion "contaminated."
Dr. Lavett believed that defendant's DNA patterns did not match the crime scene samples in this case. She did not believe it was irresponsible or unscientific of her not to have published any papers concerning her criticisms of forensic DNA testing. She admitted she received "considerably more" than $15,000 for her work on this case.
Dr. Martin Shapiro 31 is a psychologist specializing in "the interaction between quantitative methods, statistics, probability, things like that...." Although he has no specific training in human population genetics, Dr. Lavett has taught him "a lot" on the subject, and he has spent two years analyzing population data bases in DNA forensics. He has not personally created an autorad or witnessed the RFLP procedure, but he did spend three days at the FBI looking at autorads. Dr. Shapiro criticized Cellmark's "resolution unit" criteria for declaring a match, its subjectivity, and its criteria for estimating the probability of a random match. He complained that the method Cellmark used to read the autorads with a computer was an "incredibly subjective process." This was because a person operated the computer and according to Cellmark's computer manual could delete a band, move a band or separate a band. He looked at the autorads in this case and did not believe that the "pieces ... fit the puzzle."
Dr. Lyndon Larcom 32 testified that he has been researching "the conditions under which you can do electrophoresis so that the results are completely trustworthy and reproducible from one gel to the next." In this regard, he opined that Cellmark's electrophoresis experiments were deficient. Dr. Larcom believed that Cellmark's techniques were not suitable for research labs for three reasons: its protocol was not clear, its procedures as designed were defective, and the pouring of the gels is done in such a way that it does not lead to uniform results. Dr. Larcom also criticized Cellmark's use of ethidium bromide which makes the DNA fragments move slower through the gel. He also indicated there could be problems if the trays the gels were placed in were warped.
Dr. Eldred Ribeiro's 33 opinions were much like his former professor's, Dr. Larcom. He had visited the Cellmark laboratory and believed that their electrophoresis procedures were not reliable. He criticized the use of ethidium bromide, the possibility of warped gel trays, and Cellmark's resolution unit match criteria.
Dr. Mueller testified much as he did at the Kelly/ Frye hearing. He stated that he has testified for the defense in over 40 criminal cases since early 1989. It was his opinion that the scientific community was divided as to whether Cellmark's methods for estimating frequencies of a match were generally accepted. A group of scientists, including Dr. Eric Lander, Dr. Charles Taylor, and Drs. Lewontin and Hartl expressed concern about the possibility of population subgroups. Dr. Mueller observed that there is "no evidence that such subgroups don't exist, and in fact, the evidence that does exist suggests that there are subgroups within these major [racial] categories, and hence, the product rule should not be or is not an appropriate statistic to use with these sorts of data bases."
Dr. Mueller discussed the NRC report and noted that its authors included Dr. Lander, who has reservations about the forensic uses of DNA,49 34 and Dr. King, who "testifies for the prosecution." Dr. Mueller was surprised when he learned that Dr. King testifies frequently (40 percent of the time) for the defense. Using the NRC's "ceiling principle," he estimated that the likelihood of a random match of defendant's DNA pattern at the loci studied was one in 44,000. He acknowledged that none of the individuals in Cellmark's 630-person data bank matched defendant's DNA profile. By employing a different method of calculation called the "counting method" in which one looks at the number of data base allele matches in Cellmark's data base, Dr. Mueller opined that defendant's pattern "might be as common as one in 211."
Dr. Mueller was of the opinion that the frequency of a match should take into account the lab's error rate. For example, if Cellmark's error rate was one in one hundred, "the chance of a coincidental match would ... be equal to the rate at which the laboratory would make a false positive[, i.e.,] about one in a hundred." He acknowledged, however, that the NRC had proposed that the jury be given the two figures independently--both the mathematical probability based on population statistics and the lab's error rate.
Dr. Aimee Bakken, 35 who has studied human chromosomes in addition to studying frogs, sea urchins, sand dollars, zebra fish, and fruit flies, uses "methods that are used in RFLP" in her lab.
Before testifying in this case, she consulted with Dr. Lavett, Dr. Ford, and a Dr. Grunbaum, she looked at Cellmark's protocol, she read the technician's bench notes and examined the autoradiograms from this case, and she read "a number of ... publications in scientific journals and Newsweek and the New York Times that deal with various people's opinions about whether or not DNA testing is ready for the courtroom."
It was Dr. Bakken's opinion that Tony McNeil had violated Cellmark's protocols by not writing down what he was doing as he completed each step of the RFLP process. She felt that Cellmark's performance on the proficiency tests was "very disturbing," because it shows that "under these best conditions they can't come up with reliable and reproducible DNA tests...."
Dr. Bakken criticized Cellmark's methods of extracting DNA in as pure a form as possible without degradation or contamination. Another problem was that Cellmark did not insure that all the detergent (used to break down the nucleus membrane) was removed before applying the restriction enzymes. She also criticized Cellmark's reporting of a match even though extra bands were present. She felt the proficiency test autorads demonstrated Cellmark's carelessness.
III
DISCUSSION
A. DNA ISSUES
Defendant contends the trial court erred in failing "to hold a Kelly/ Frye hearing at all"; "Axell has been undermined by 'a change in the scientific community' "; Cellmark's "match" procedures and statistical analysis have not achieved general acceptance in the relevant scientific community; and Cellmark did not follow the proper procedures in this case.
1. Standard of Review
Review of a trial court's decision finding that a new scientific procedure has been "sufficiently established to have gained general acceptance in the particular field in which it belongs" (internal quotes omitted) and therefore that it is admissible in a criminal trial is a "mixed question of law and fact subject to limited de novo review." (People v. Reilly (1987) 196 Cal.App.3d 1127, 1133-1134, 242 Cal.Rptr. 496.)
The appellate court reviews "the trial court's determination with deference to any and all supportable findings of 'historical' fact or credibility, and then decide[s] as a matter of law, based on those assumptions" not " 'whether [the technique] is reliable as a matter of "scientific fact," but simply whether it is generally accepted as reliable by the relevant scientific community.' [Citations.]" (People v. Reilly, supra, 196 Cal.App.3d at p. 1135 and fn. 1, 242 Cal.Rptr. 496.)
In conducting the review, the appellate court primarily relies on the record below (People v. Reilly, supra, 196 Cal.App.3d at pp. 1134-1135, 242 Cal.Rptr. 496; People v. Morris (1988) 199 Cal.App.3d 377, 387, 245 Cal.Rptr. 52) but it may also consider out of state opinions and scientific literature bearing on the question whether the scientific technique at issue has gained general acceptance in the scientific community. (People v. Shirley (1982) 31 Cal.3d 18, 56, 181 Cal.Rptr. 243, 723 P.2d 1354; People v. Leahy, supra, 8 Cal.4th at p. 611, 34 Cal.Rptr.2d 663, 882 P.2d 321.)
2. Failure to Hold a Kelly/Frye Hearing
Notwithstanding defendant's contention that no Kelly/ Frye hearing was held here "at all," an extensive hearing took place in August and November 1989 at which six prosecution and four defense expert witnesses opined on whether Cellmark's RFLP process and its statistical analysis were generally accepted in the scientific community. (Testimony from that hearing is summarized ante in Background section D.) In addition, counsel for both sides provided the magistrate with numerous scientific articles on DNA typing. After considering the evidence, the magistrate admitted the "results of the sperm DNA matching on both Lisa [K.] and Martha [D.], and the blood DNA given by the Defendant Marlow" at defendant's preliminary hearing.
Defendant's actual complaint is that a second Kelly/ Frye hearing was not held by the trial court before admitting DNA evidence at trial. However, here too defendant's argument is flawed. As noted earlier, the trial court agreed to spend the time necessary to read the transcript from the first Kelly/ Frye hearing. 36 Throughout the numerous discovery and discovery compliance hearings, the court repeatedly indicated its intent to gain additional knowledge on DNA testing in order to clarify and narrow the issues to be decided in a Kelly/ Frye hearing.
In furtherance of this goal, the court allowed two defense expert witnesses to testify about Cellmark's DNA RFLP testing procedures at one of the discovery motion hearings. Dr. Eldred Ribeiro testified that during his last visit to the Cellmark laboratory, he had "observed certain deficiencies and inadequacies in three important areas of the profiling process: the first being the DNA sample preparations, second the electrophoresis, and third the analysis of the autoradiographs that are made from the [S]outhern blots, which in turn are made from these gels." Dr. Ribeiro found fault with Cellmark's use of ethidium bromide "to visualize the DNA and follow it during the electrophoresis process," its use of allegedly warped trays (in which the gel is placed), its failure to follow its own protocol, and its procedures for declaring a match.
The court also permitted Dr. Paul Hagerman 37 to testify. Like Dr. Ribeiro, Dr. Hagerman did not approve of Cellmark's use of ethidium bromide. He believed that placing ethidium bromide in the gels before electrophoresis could cause "significant band shifting." In addition, he criticized Cellmark's resolution unit method for comparing band sizes to declare a match. He said Cellmark's resolution unit concept was not generally accepted in the scientific community.
On October 29, 1991, the Axell decision was filed. (People v. Axell, supra, 235 Cal.App.3d 836, 1 Cal.Rptr.2d 411.) As noted earlier, the Axell court held that the RFLP procedure used by Cellmark was one that was generally accepted as reliable in the scientific community. The court also held that since a "match" between two DNA samples means little without data on the match's statistical significance, the calculation for statistical probability must also pass Kelly/ Frye muster; i.e., it must meet the same test for admissibility as the evidence of the RFLP technology itself. Finally, the Axell court held that the method Cellmark employs to arrive at a statistical probability is a method generally accepted in the scientific community.
At this point, the district attorney filed a motion to admit DNA results without further hearing based upon the Axell decision and upon the California Supreme Court's statement in People v. Kelly, supra, 17 Cal.3d at p. 32, 130 Cal.Rptr. 144, 549 P.2d 1240, that "once a trial court has admitted evidence based upon a new scientific technique, and that decision is affirmed on appeal by a published appellate decision, the precedent so established may control subsequent trials, at least until new evidence is presented reflecting a change in the attitude of the scientific community." (See also People v. Leahy, supra, 8 Cal.4th at p. 595, 34 Cal.Rptr.2d 663, 882 P.2d 321.)
Defendant requested an opportunity to present evidence of a change in the consensus within the scientific community. The court granted the request but indicated that the evidence would have to be presented by way of declarations. On March 31, 1992, defendant filed his response, including 22 exhibits, and totaling 539 pages in length. Defendant argued there were "two major concerns" voiced by critics of DNA forensics, to wit: "First, they believe that the procedures used for determining whether two DNA prints match or do not match are inadequately validated and unreliable. Second, they believe that the procedures used to determine the statistical rarity of DNA prints and the probability of a coincidental match are invalid and unreliable." Eleven scientific articles critical of some aspect of DNA testing were appended, including Population Genetics in Forensic DNA Typing, 254 Science 1745 (December 20, 1991) by Lewontin and Hartl (hereafter Lewontin and Hartl); and Fight Erupts Over DNA Fingerprinting, 254 Science 1721 (December 20, 1991). Defendant also included a survey of 30 population geneticists, 19 of whom expressed criticism with some aspect of DNA testing. Defendant pointed out there were new questions about the use of ethidium bromide, rater bias, DNA concentration effects, and population subgrouping. Also included were declarations criticizing Cellmark's match procedures, statistical analysis, computer matching system, and data bases.
After considering the Kelly/ Frye testimony below (almost 2000 pages), the additional testimony taken in connection with a discovery motion (approximately 135 pages), the 539 pages of scientific articles and declarations submitted by defendant, the Axell opinion, and the Supreme Court's statement that Kelly/ Frye hearings need not be endlessly repeated if a particular new scientific technique has been upheld on appeal and there does not appear to be a change in the scientific consensus, the court determined to admit DNA evidence without a further Kelly/ Frye hearing.
On appeal, defendant contends this was error. We disagree. As the chronology of this case amply demonstrates, the trial court did not blindly accept the Axell holding. Rather, it educated itself about DNA RFLP testing. It read the transcript of the Kelly/ Frye hearing below and analyzed which experts seemed most knowledgeable. As the California Supreme Court recently instructed: "[T]he trial courts, in determining the general acceptance issue, must consider the quality, as well as quantity, of the evidence supporting or opposing a new scientific technique. Mere numerical majority support or opposition by persons minimally qualified to state an authoritative opinion is of little value under the foregoing cases." (People v. Leahy, supra, 8 Cal.4th at p. 612, 34 Cal.Rptr.2d 663, 882 P.2d 321.) Here, all the prosecution experts worked with the RFLP process on a daily basis. In contrast, the defense experts' knowledge of the subject appeared to come primarily from testifying for the defense in various cases. None had the "hands-on" familiarity with the process that the prosecution experts had. Defense expert Dr. Lavett's credibility has been challenged in other published cases. (See e.g., People v. Reilly, supra, 196 Cal.App.3d at p. 1151, fn. 7, 242 Cal.Rptr. 496 [noting her lack of "practical experience" and finding "her added voice [did not tip] the Kelly/ Frye scale in defendant's favor"]; People v. Axell, supra, 235 Cal.App.3d at p. 859, 1 Cal.Rptr.2d 411 [noting she was not "particularly credible" as she had not done RFLP analysis herself].)
After considering the evidence, the court made an informed decision that Axell had correctly found that Cellmark's RFLP testing was generally accepted in the relevant scientific community and that although there were a few dissenting voices, 38 there had not been a change in the general acceptance of the RFLP procedure by the relevant scientific community since Axell.
We find no fault with the court's decision not to hold another Kelly/ Frye hearing. The avoidance of repetitive hearings after a procedure has once been found to be generally accepted in the scientific community is precisely what was contemplated in the Kelly opinion.
3. Have the "Match" and Statistical Analysis Procedures that Were Found to Pass Kelly/Frye Muster in Axell Been Undermined by Subsequent Changes in the Scientific Community?
After the trial court decided to admit DNA evidence without further Kelly/ Frye hearings, two events occurred which defendant deems significant. First, on April 16, 1992, just days before defendant's trial began, the National Research Council issued its 185 page report, DNA Technology in Forensic Science (hereafter NRC report). 39 Second, on August 5, 1992, shortly after the verdict was reached here, Division Three of the First District issued its People v. Barney opinion, supra, 8 Cal.App.4th 798, 10 Cal.Rptr.2d 731. 40
The NRC report was the culmination of three years of study by a prominent group of scientists, lawyers, and ethicists, including prosecution witness Dr. Mary-Claire King. The goal of the committee was to establish a national policy regarding the forensic use of DNA technology, to develop guidelines and safeguards, and to provide recommendations for legislation.
In discussing whether DNA typing evidence satisfied the admissibility standards of Frye, the NRC report noted: (1) "[T]hat ... each person's DNA is unique [ ] is so well established in human molecular genetics that a court is justified in judicially noticing it"; (2) "[T]he validity of procedures for extracting DNA from samples of blood, semen, and other materials and analyzing it for the presence and size of polymorphisms ... is sufficiently well established in the case of RFLP analysis with Southern blots that judicial notice is also appropriate"; 41 (3) "[T]he adequacy of statistical databanks used to calculate match probabilities ... rests on unproven foundations.... The solution, however, is not to bar DNA evidence, but to ensure that estimates of the probability that a match between a person's DNA and evidence DNA could occur by chance are appropriately conservative" "; (4) "[T]hat the analytical work done for a particular trial comports with proper procedure ... can be resolved only case by case and is always open to question, even if the general reliability of DNA typing is fully accepted in the scientific community." (NRC Report, pp. 133-134; emphasis added.)
With respect to the statistical basis for interpreting matches, the NRC report stated that "[t]o say that two patterns match, without providing any scientifically valid estimate (or, at least, an upper bound) of the frequency with which such matches might occur by chance, is meaningless." 42 (NRC Report p. 74.)
The report observed that "[s]ubstantial controversy has arisen concerning the methods for estimating the population frequencies of specific DNA typing patterns. Questions have been raised about the adequacy of the population databases on which frequency estimates are based...." (NRC report at pp. 74-75.) The "product rule" or "multiplication rule" 43 used by forensic laboratories to estimate the frequency of finding another individual whose DNA pattern matches on all of the loci studied assumes that those populations are in Hardy-Weinberg equilibrium 44 and in linkage equilibrium. 45 If these assumptions are not correct (as is theorized in the case of population subgroups), the statistical analysis may be incorrect. "The key question underlying the use of the multiplication rule is whether actual populations have significant substructure for the loci used for forensic typing." (Id. at p. 79.)
The NRC committee "decided to assume that population substructure might exist for currently used DNA markers...." (NRC Report, p. 94; emphasis added.) Accordingly, to compensate for any possible substructuring, it recommended use of a modified product rule known as the "ceiling principle." 46 The committee noted that the ceiling principle would "yield conservative estimates, even for a substructured population, provided that the allele frequencies used in the calculation exceed the allele frequencies in any of the population subgroups." (Id. at p. 82; emphasis added.) Under the ceiling principle, "[r]andom samples of 100 persons should be drawn from each of 15-20 populations, each representing a group relatively homogeneous genetically; the largest frequency in any of these populations or 5%, whichever is larger, should be taken as the ceiling frequency." (Id. at p. 83.) Then the resulting frequencies may be multiplied together to obtain a final ceiling frequency calculation. Until such time as the above recommended studies of 15 to 20 population groups are undertaken and studied, an interim approach, known as the "modified ceiling principle" should be employed. Under this approach, laboratories should find the most common frequency from the existing data banks but use a higher bound of 10 percent (or the actual frequency, if larger). (Id. at p. 92.)
The second major event undermining Axell, according to defendant, was the August 1992 publication of People v. Barney, supra, 8 Cal.App.4th 798, 10 Cal.Rptr.2d 731. In Barney, the First District Court of Appeal, which purported to rely heavily on the NRC report, essentially ignored the conclusion of that committee, i.e., that "[t]he solution [to the question of substructuring] is not to bar DNA evidence, but to ensure that estimates of the probability that a match between a person's DNA and evidence DNA could occur by chance are appropriately conservative." (NRC Report p. 134.) Relying on the NRC report's statement that a "[s]ubstantial controversy" exists concerning the present method of statistical analysis, and on Lewontin and Hartl's article Population Genetics in Forensic DNA Typing, 254 Science 1745 supra, the court held that DNA evidence was not admissible under Kelly/ Frye because scientists were not in agreement on the statistical analysis. 47 (Id. at p. 819, 10 Cal.Rptr.2d 731.)
Under the heading "General Acceptance and Statistical Analysis: [p] The current scientific debate," the court devoted four paragraphs to summarizing Lewontin and Hartl's contentions (that subgroups may have substantial differences in the frequency of a given DNA fragment and that the current product rule may yield errors of up to two orders of magnitude if populations are not in Hardy-Weinberg equilibrium and linkage equilibrium). It followed this lengthy discussion with a simple observation that Chakraborty and Kidd, who wrote a companion article (The Utility of DNA Typing in Forensic Work ) in the same December 20, 1991, issue of Science, "strongly disagree." (People v. Barney, supra, 8 Cal.App.4th at pp. 814-815, 10 Cal.Rptr.2d 731.)
Later, under the heading "General Acceptance and Statistical Analysis [p] 4. General scientific acceptance," the court devoted two more paragraphs to this single article by Lewontin and Hartl. The court noted that others were also critical of the statistical calculation process of DNA analysis, including two whose articles had been admitted into evidence below (Lander, DNA Fingerprinting on Trial (June 15, 1989) Nature pp. 501, 504 48 and Cohen, DNA Fingerprinting for Forensic Identification: Potential Effects on Data Interpretation of Subpopulation Heterogeneity and Band Number Variability (1990) 46 Am.J.Hum. Genetics pp. 358, 367). Again, however, the court neglected Lewontin and Hartl's opponents. The court's discussion of the opponents' position reads, in full: "Of course, Chakraborty and Kidd strongly disagree, and according to Science they have 'many scientific supporters.' (Fight Erupts, supra, at p. 1721; see Risch & Devlin On the Probability of Matching DNA Fingerprints (Feb. 7, 1992) Science, at p. 717.)" (People v. Barney, supra, 8 Cal.App.4th. at p. 819, 10 Cal.Rptr.2d 731.)
As we shall explain, we believe Barney's analysis was inadequate. First, the court appears to have been unduly influenced by a single article by Lewontin and Hartl. Indeed, no other scientific point of view is discussed or analyzed. None of the expert witnesses' testimony from the Kelly/ Frye hearing appears in the opinion and it appears to have been ignored by the court in making its conclusion. The court also appears to have placed great reliance on statements in secondary sources reporting on the rift between the Lewontin and Hartl camp and other forensic scientists.
We believe that in order to ascertain whether a certain procedure has gained general acceptance in the scientific community, it is incumbent to look first to the testimony of the experts who testified as to the new scientific procedure and second to primary sources (i.e., scientific articles). If these reveal that a consensus drawn from a typical cross-section of the relevant, qualified scientific community accepts the procedure as reliable, it will pass muster under Kelly. (See generally, People v. Leahy, supra, 8 Cal.4th at p. 611-612, 34 Cal.Rptr.2d 663, 882 P.2d 321.)
We also take issue with the "head counting" approach apparently employed by the Barney court. In deciding the general acceptance issue, the trial court (and the appellate court on de novo review) must not simply count heads but must look to the quality as well as the quantity of evidence supporting or opposing a new scientific method. As the California Supreme Court explained in Leahy, " 'resolution of the general acceptance issue ... require[s] consideration of the views of a typical cross-section of the scientific community....' " (People v. Leahy, supra, 8 Cal.4th at p. 611, 34 Cal.Rptr.2d 663, 882 P.2d 321.) Citing its earlier opinion in People v. Guerra (1984) 37 Cal.3d 385, 208 Cal.Rptr. 162, 690 P.2d 635, the court explained that Frye does not demand " 'absolute unanimity of views in the scientific community....' " (Id. at p. 612, 34 Cal.Rptr.2d 663, 882 P.2d 321.) The court advised "trial courts, in determining the general acceptance issue, [to] consider the quality, as well as quantity, of the evidence supporting or opposing a new scientific technique." (Ibid.) It pointed out that "[m]ere numerical majority support or opposition by persons minimally qualified to state an authoritative opinion is of little value under the foregoing cases." (Ibid.)
As we shall explain, we conclude that a "typical cross-section of the relevant, qualified scientific community" accepts DNA RFLP analysis (including the determination of a match and the statistical interpretation of the match by use of the product rule) as reliable. 49 The testimony of the experts who testified in this case--at the Kelly/ Frye hearing, at the discovery motion, and at trial--is recounted in detail supra and will not be repeated here. However, when one examines the relative qualifications of the experts, their actual experience with RFLP, their knowledge of human population genetics (as compared, for example, with fruit fly genetics), and their financial interest in testifying, it is clear that the experts supporting RFLP analysis were more credible. This was the conclusion of the trial court who heard all the testimony, and this is our conclusion on de novo review.
Furthermore, our analysis of a broad cross-section of the scientific literature discloses a general acceptance of RFLP by the relevant, qualified scientific community. 50 We shall begin our review of the literature with Lewontin and Hartl, since that is where Barney began. Those two clearly prominent scientists (the former from Harvard, the latter from Washington University) first raised the issue of the possible effect of substructuring on DNA probability estimates when they testified for the defense at a six-week long Frye hearing in the murder prosecution of defendants Yee, Verdi and Bonds. 51 (U.S. v. Yee (N.D.Ohio 1991) 134 F.R.D. 161 [defendant Bonds' DNA was found in the victim's car and in Yee's get-away car]; affd. U.S. v. Bonds, supra, 12 F.3d 540.) Lewontin and Hartl complained that "the possibility of substructuring invalidates the process of multiplication to obtain a probability estimate" because the frequency of DNA bands might vary among different ethnic groups.
Lewontin next testified in U.S. v. Jakobetz (D.Vt.1990) 747 F.Supp. 250; affd. U.S. v. Jakobetz (2d Cir.1992) 955 F.2d 786, explaining that "[e]vidence of substructure ... undermine[s] [the] assumption of independence" necessary to allow application of the product rule. (U.S. v. Jakobetz, supra, 955 F.2d at p. 799.) Among those testifying for the prosecution in U.S. v. Jakobetz, supra, were Dr. Kidd (who co-authored The Utility of DNA Typing in Forensic Work, supra, with Dr. Chakraborty), Dr. Conneally (who testified in this case) 52 and Dr. Budowle (who co-authored the article with Dr. Eric Lander [see fn. 34] declaring the DNA wars to be over).
The reports Lewontin and Hartl used in Yee and the theories espoused in Jakobetz formed the substance of the December 20, 1991, Science article relied upon so heavily by the Barney court. Just two months after that article was published, Science editor Daniel E. Koshland, Jr. explained the genesis of that article. He noted that numerous "peer reviewers" and "several [other] geneticists at the October 1991 International Congress of Human Genetics" had advised him that "some of Lewontin's and Hartl's more theoretical arguments were not adequately supported by data." (Koshland, Letters, 255 Science (Feb. 28, 1992), p. 1053.) Because of this criticism, Koshland asked Lewontin and Hartl to submit their article as an "opinion piece" rather than as a scientific article. When Lewontin and Hartl refused, Science decided to go ahead and publish their article, but to also publish an accompanying article pointing out the flaws. (This was the article, barely discussed by Barney, by Chakraborty and Kidd, The Utility of DNA Typing in Forensic Work, supra, 254 Science pp. 1735-1739.)
As noted earlier, despite Lewontin's and Hartl's testimony in U.S. v. Yee, supra, and Lewontin's testimony in U.S. v. Jakobetz, supra, the magistrates in both cases found the RFLP test results, including the population frequency estimates based on the product rule, to be admissible under Frye. In other words, the magistrates found that the theories advanced by Lewontin and Hartl were not those that represented the consensus view in the relevant scientific community. In each case, the magistrate's findings were upheld by the federal district court, which admitted the RFLP test results into evidence. In each case, the defendants were convicted. Each conviction was in turn affirmed on appeal. (U.S. v. Bonds, supra, 12 F.3d 540; U.S. v. Jakobetz, supra, 955 F.2d 786.)
In U.S. v. Bonds, the Sixth Circuit explained that complete unanimity is not required when determining whether a new procedure has gained the general acceptance of the scientific community. The court stated, "The evidence and testimony presented at this Frye hearing demonstrated that the DNA evidence was not based on untested or unacceptable theories or procedures. Because the DNA results were based on scientifically valid principles and derived from scientifically valid procedures, it is not dispositive that there are scientists who vigorously argue that the probability estimates are not accurate or reliable because of the possibility of ethnic substructure. The potential of ethnic substructure does not mean that the theory and procedures used by the FBI are not generally accepted; it means only that there is a dispute over whether the results are as accurate as they might be and what, if any, weight the jury should give those results." (12 F.3d at pp. 564-565.)
Furthermore, since Lewontin's and Hartl's article appeared in Science, numerous studies have been published in scientific journals compiling VNTR frequency data from around the world. These studies have empirically demonstrated the very conservative nature of the frequency calculation methods employed by forensic laboratories, i.e., the product rule and its variation, the ceiling principle. (See e.g., U.S. Department of Justice, VNTR Population Data: A Worldwide Study, Vols. I-A, (1993) [After compiling extensive empirical data, the study concludes "(1) that there are sufficient population data available to determine whether or not forensically significant differences might occur when using different population data bases; (2) that subdivision, either by ethnic group or by U.S. geographic region, within a major population group does not substantially affect forensic estimates of the likelihood of occurrence of a DNA profile; (3) that estimates of the likelihood of occurrence of a DNA profile using major population group databases (e.g., Caucasian, Black, and Hispanic) provide a greater range of frequencies than would estimates from subgroups of a major population category; therefore, the estimate of the likelihood of occurrence of a DNA profile derived by the current practice of employing the multiplication rule and using general population databases for allele frequencies is reliable, valid, and meaningful, without forensically significant consequences; and (4) that the data do not support the need for alternate procedures, such as the ceiling principle approach (NRC Report 1992), for deriving statistical estimates of DNA profile frequencies."]; Budowle, Monson, Giusti, and Brown, The Assessment of Frequency Estimates of Hae III-Generated VNTR profiles in Various Reference Databases, 39 Journal of Forensic Sciences, No. 2, (March 1994) pp. 319-352; Budowle, Monson, Giusti, and Brown, Evaluation of Hinf I-Generated VNTR Profile Frequencies Determined Using Various Ethnic Databases, 39 Journal of Forensic Sciences, No. 4 (July 1994) pp. 988-1008; Chow, Tan, Yap, Ng, The Development of DNA Profiling Database in an HAE III Based RFLP System for Chinese, Malays, and Indians in Singapore," 38 Journal of Forensic Sciences, No. 4 (July 1993) pp. 874-884 ["[c]omparison of the ratios of bin frequencies [from Chinese, Malay, and Indian population data with published data on Whites, Blacks, and Hispanics] shows that difference of allelic distribution at these loci is not forensically significant"].)
Moreover, the weight of authority in the published peer-reviewed literature overwhelmingly supports the proposition that VNTR frequency differences due to ethnicity or substructuring have little impact on DNA population frequency estimates and that the product rule is appropriate to estimate probabilities of a random match. (See e.g., Chakraborty and Kidd, The Utility of DNA Typing in Forensic Work, supra; Risch and Devlin, On the Probability of Matching DNA Fingerprints, 255 Science (February 7, 1992) 717-720 [analyzed FBI's and Lifecode's data bases]; Devlin and Risch, Ethnic Differentiation at VNTR Loci, with Special Reference to Forensic Applications, 51 Am.J.Hum.Genet. (1992) 534-548 [alleles common in one group almost always common in other groups]; Devlin and Risch, A Note on Hardy-Weinberg Equilibrium of VNTR Data by Using the Federal Bureau of Investigation's Fixed Bin Method, 4 Am.J.Hum.Genet. (1992) 549-553; Chakraborty and Daiger, Polymorphism at VNTR Loci Suggest Homogeneity of the White Population of Utah, 63 Human Biology, No. 5, (October 1991) pp. 571-587; Chakraborty, Deka, Jin, Ferrell, Allele Sharing at Six VNTR Loci and Genetic Distances Among Three Ethnically Defined Human Populations, 51 Am.J.Hum.Biology (1992) 387-397; Chakraborty and Jin, Heterozygote Deficiency, Population Substructure and Their Implications in DNA Fingerprinting (1992) 88 Hum.Genet. 267-272 ["[T]he possible effect of realistic substructuring that exists within major racial groups of humans is almost negligible...."]; Devlin, Risch, and Roeder, Statistical Evaluation of DNA Fingerprinting: A Critique of the NRC's Report (February 5, 1993) 259 Science 748-749, 837; Anderson, Courts Reject DNA Fingerprinting, Citing Controversy After [NRC] Report (October 1, 1992) 359 Nature 349 ["DNA evidence continues to be held to a higher standard than other scientific subjects. That standard, under Frye, requires only 'general acceptance' in the scientific community. Yet DNA evidence is often being rejected on the basis of controversies involving a few scientists, including Lewontin, who proclaim themselves to be extremists."]; Chakraborty, Letter (May 21, 1993) 260 Science 1059; Herrin, Probability of Matching RFLP Patterns from Unrelated Individuals, 52 Am.J.Hum.Genet. (1993) 491-497; Herrin, Technical Note, A Comparison of Models Used for Calculation of RFLP Pattern Frequencies (Nov. 1992) 37 Journal of Forensic Sciences No. 6, 1640-1651.)
Based on this review of the literature, we conclude that a typical cross-section of the relevant, qualified scientific community accepts as reliable the frequency estimate methods employed by Cellmark, i.e., use of the product rule and population data bases to estimate the likelihood of finding a matching DNA profile. 53
We also conclude that Cellmark's procedure for declaring a match 54 is generally accepted in the scientific community, notwithstanding defendant's experts' complaints about Cellmark's resolution unit methodology. As we have noted earlier, once a court has found a certain new scientific technique to have passed Kelly/ Frye muster, and once that opinion is affirmed in a published opinion on appeal, the issue need not be endlessly relitigated. In California, every court that has analyzed Cellmark 's method of declaring a match has found it to be generally accepted as reliable in the relevant, qualified scientific community. (See e.g., People v. Axell, supra, 235 Cal.App.3d at p. 860-863, 1 Cal.Rptr.2d 411; People v. Barney, supra, 8 Cal.App.4th at pp. 812-814, 10 Cal.Rptr.2d 731.) 55 In addition, the NRC report concluded that "in the case of RFLP analysis," a court could simply take judicial notice of the validity of the "procedures for extracting DNA from samples of blood, semen, and other materials and analyzing it for the presence and size of polymorphisms" (i.e., measuring the bands to declare a match).
Accordingly, we find that there has not been a change in the scientific consensus since Axell which in any way undermines that decision.
4. Were the Proper Procedures Followed in this Case?
Defendant claims the defense experts identified a number of specific problems with Cellmark's laboratory procedures and demonstrated how those problems infected the work done in this case. Most of these "problems" concern Cellmark's general methodology, however, which we held in section 3, ante, to be generally accepted in the scientific community. Only a few of the "problems" deal with the specifics of this case.
The "problems" defendant has identified include: (1) Cellmark's protocol (not "drafted in a way that any technician could follow the Cellmark procedures"); (2) buffer concentrations (possibility of too much salt could cause the DNA to migrate across the gel more slowly); (3) failure to recirculate buffer, which affects pH level; (4) unlevel gel trays; (5) failing to cool the gels in a manner leading to gel uniformity; (6) extraction procedures (do not necessarily remove all the detergent that is used to break down the cell membrane); (7) restriction enzymes (may have been stored in too high a concentration of glycerol); (8) failure to preheat kilobase markers; (9) poor quality autorads; (10) use of ethidium bromide or using too high a concentration of DNA (both of which can cause band shifting); and (11) Cellmark's protocol for reading autorads is deficient.
Of these, only three relate to specifics in this case. First, Dr. Lavett complained that the crime scene samples were aged and contaminated while defendant's samples, coming from a blood sample, were sterile and clean. However, as Drs. King, Kovacs, Cotton, and Bowcock explained, where a sample is contaminated, no DNA pattern will emerge. Getting a false positive, according to these experts, is not a possibility. Here, a DNA pattern did emerge from the crime scene evidence, and that pattern matched defendant's. Thus, contamination did not affect the results in this case. Second, technician Tony McNeil did not keep complete notes on whether the suspect's DNA was processed with at least 10 percent glycerol. However, as Dr. Cotton testified, McNeil followed Cellmark's protocols in this case, and all his measurements were gone over by two Ph.D.'s. Finally, defendant complained that the gel tray used to run his sample may have been warped or tilted. However, as noted earlier, Dr. Conneally testified this issue was a red herring: "Well, you could tilt it this way and that way and every way in the world and keep tilting it for the whole number of hours that this was [run], and it wouldn't make any difference."
We conclude that the proper procedures were followed in this case. Accordingly, the court did not err when it admitted into evidence the DNA RFLP test results in this case. B. Non-DNA Issues **
DISPOSITION
The judgment is affirmed.
WUNDERLICH and MIHARA, JJ., concur. --------------- * Pursuant to California Rules of Court, rule 976.1, this opinion is certified for publication with the exception of section IIIB. 1 See generally, People v. Kelly (1976) 17 Cal.3d 24, 130 Cal.Rptr. 144, 549 P.2d 1240; People v. Leahy (1994) 8 Cal.4th 587, 604, 34 Cal.Rptr.2d 663, 882 P.2d 321. 2 A Pony company employee testified that 52,000 pairs of Pony shoes sold in the United States between 1986 and 1987 had the same tread pattern. 3 Evidence of the December 16, 1987, Amy L. rape was not admitted in this trial. Amy had identified defendant as her "possible" rapist, and DNA testing of defendant's blood matched the semen found in Amy. 4 Evidence of the December 22, 1986, Karen C. attack was also not admitted in this trial. We affirmed defendant's commitment to the California Youth Authority (CYA) in connection with the Karen C. incident in In re Gustavo M. (1989) 214 Cal.App.3d 1485, 263 Cal.Rptr. 328. 5 See People v. Kelly, supra, 17 Cal.3d 24, 130 Cal.Rptr. 144, 549 P.2d 1240, a California case following the rule enunciated in Frye v. United States (D.C.Cir.1923) 293 F. 1013, that before the results of a new scientific technique can be admitted in court, it must be shown that the technique has gained general acceptance in the relevant scientific community. In Kelly, the California Supreme Court explained that "[a]dmissibility of expert testimony based upon the application of a new scientific technique traditionally involves a two-step process: (1) the reliability of the method must be established, usually by expert testimony, and (2) the witness furnishing such testimony must be properly qualified as an expert to give an opinion on the subject. [Citations.] Additionally, the proponent of the evidence must demonstrate that correct scientific procedures were used in the particular case. [Citations.]" (Id. at p. 30, 130 Cal.Rptr. 144, 549 P.2d 1240, italics in original omitted.) Kelly further held that once the technique had been found to be generally accepted in the relevant scientific community, it would not be necessary to have repetitive hearings on the same technique. (Id. at p. 32, 130 Cal.Rptr. 144, 549 P.2d 1240.) In People v. Leahy, supra, 8 Cal.4th at page 591, 34 Cal.Rptr.2d 663, 882 P.2d 321, the California Supreme Court reaffirmed the principles enunciated in Kelly, notwithstanding the United States Supreme Court's abrogation of Frye in Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) 509 U.S. 579, ----, 113 S.Ct. 2786, 2794, 125 L.Ed.2d 469. 6 These included: (1) the Amy L. incident (see fn. 3); (2) the Karen C. incident (see fn. 4); (3) a July 23, 1988, incident involving a victim who escaped her attacker by running into a pizza parlor; and (4) a June 11, 1991, incident on the grounds of the California Youth Authority in which an inmate, who was positively identified as defendant, beat, raped, gagged, and robbed a CYA employee and then escaped in her car. 7 People v. Marsden (1970) 2 Cal.3d 118, 123, 84 Cal.Rptr. 156, 465 P.2d 44. 8 The only exception is egg and sperm cells, which have 23 individual chromosomes. 9 Again, except for egg and sperm cells, which have half the complement of DNA present in other human cells. However, one may determine a man's complete DNA profile from a sperm sample because it contains thousands of DNA molecules and thousands of sperm. 10 In contrast, the "coding" parts of the DNA, the genes, are the fundamental units of heredity and perform biological functions. 11 For example, "one VNTR in humans consists of a 17-base pair unit repeated from 70 to 450 times at one locus in the genome. Thus, the total number of base pairs would vary from 1,190 (17 X 70) to 7,650 (17 X 450)." (U.S. Congress, Office of Technology Assessment, Genetic Witness: Forensic Uses of DNA Tests, (1990) p. 44 [hereafter OTA report].) 12 The genome consists of all the genetic material in the chromosomes of a particular organism. Its size is generally given as its total number of base pairs. 13 Dr. Southern is the scientist who developed this procedure. 14 Dr. Bowcock received her Ph.D. from the University of South Africa. 15 These are DNA fragments of a known base pair length which are run on the gels. The sample DNA fragments can then be measured against them, like a ruler. 16 He received his medical degree from the University of Southern California (USC) and specializes in maternal fetal medicine and human genetics. He currently is director of USC's molecular and science laboratory at its medical school. 17 Dr. Cotton is Cellmark's research and development manager. She received her Masters of Science degree from Methodist University in Texas, her Ph.D. in molecular biochemistry and molecular biology from the University of California at Irvine, and did post doctoral research at the University of Iowa. Prior to coming to Cellmark, Dr. Cotton worked at the National Institute of Health in Maryland investigating the role of genetics in alcoholism. 18 She received her bachelor of science degree from the University of California and her master's and Ph.D. in biological anthropology (studying population genetics) from New York University. She did research for the Smithsonian Institute on genetic profiles of endangered species, and had authored numerous scientific articles including four on the forensic use of DNA. 19 This is referred to as the "bin size." Bins lump together similar measurements. For example, a "bin" might include all individuals who are between 5 feet 7 1/2 inches and 5 feet 8 1/2 inches. Such a "bin" would obviously be much more common than a bin limited to only those measuring exactly 5 feet 8 and 3/16 inches. 20 She received her bachelor's degree from Carleton College and her Ph.D. in biogenetics from the University of California, where she is now a professor of epidemiology. At the behest of Amnesty International, which is attempting to reunite children whose parents were killed during military rule in Argentina with their living grandparents, Dr. King has done DNA testing of the children and potential grandparents to establish a familial DNA link. Dr. King has published close to 100 papers and has testified in numerous cases. 21 Dr. Lavett completed her Ph.D. and post doctoral research in genetics at Emory University. She is an associate professor at the State University of New York at Cortland and is the author of the "Student Companion with Complete Solutions for an Introduction to Genetic Analysis." 22 He obtained his bachelor's degree in genetics at the University of Leeds and his Ph.D. in biochemistry from the University of Bristol. He is a research specialist at the University of California, Irvine. 23 Dr. Thompson is an associate professor at the University of California, Irvine. He received his J.D. from the University of California and his Ph.D. in psychology from Stanford. 24 The complete colloquy between judge and expert is reprinted here to elucidate how the two arrived as such different conclusions from the same survey results: "[THE WITNESS]: They were given four options, and they split between two of the options: 'The test is definitely reliable, I'd bet my life on it.' " "THE COURT: I'm showing you the cover page on Defendant's 83, 'Survey of Molecular Biologists Reliability of Lifecodes and Cellmark.' I don't want to beat a dead horse, but this is the one Mr. Damkar asked you about, would you bet your life on the reliability. [p] THE WITNESS: Yes. [p] THE COURT: A group answered, definitely reliable, they would bet their life; a group answered, I wouldn't bet my life, but it's probably reliable. [p] THE WITNESS: Right. [p] THE COURT: And no one said it was not reliable? [p] THE WITNESS: That's right. [p] THE COURT: Now, what conclusion do you reach, either favorable or unfavorable, as to the reliability of Cellmark's test? [p] THE WITNESS: Among those who know something about the tests, who are a small portion of the total sample, there's a split between those who think it's reliable enough to bet their life on it and those who think it's probably reliable, but aren't sure it's that--reliable enough to bet their life on it. [p] THE COURT: So the--excuse the expression--the community that answered the question, answered it that it was reliable? [p] THE WITNESS: Um-hum. [p] THE COURT: And no one said it was not reliable? [p] THE WITNESS: Right. [p] THE COURT: Now, would you conclude that that would be favorable to the Cellmark tests--the reliability of the Cellmark tests? [p] THE WITNESS: Somewhat. You know, again, it gets back to the use of this term 'reliable,' and the different--the different meanings placed on that term, and the different meanings placed on 'reliability,' depending on context. [p] THE COURT: Everyone who answered said it was reliable. [p] THE WITNESS: That's true; no question about it. [p] THE COURT: Now, if everyone says it's reliable that answered, wouldn't you consider that more favorable than anyone--than if everyone replied that it was not reliable? [p] THE WITNESS: No doubt about it. [p] THE COURT: No doubt about it? [p] THE WITNESS: Sure. [p] THE COURT: But you will not concede that this has any bearing on the reliability of Cellmark's testing? [p] THE WITNESS: Oh, no, I wouldn't. That's not what I'm saying. [p] THE COURT: What are you saying? [p] THE WITNESS: Certainly, this is relevant, something that should be taken into account, that everyone who, you know, thought they knew something about Lifecodes and Cellmark said it was either definitely reliable or probably reliable. [p] What concerns me is that, given the--given that this is a test that's going to be used such that very serious decisions are based on a single result, that there are quite a few who said, although probably reliable, it's not something they would want to bet their life on, so really, what we're dealing with here, I think, is a question of what level of reliability is required for Frye purposes. [p] THE COURT: That's why we're here. Frye doesn't say bet-your-life reliability, Frye indicates, since 1923, and most of the states that have followed it indicate that it has to be generally reliable. Now that doesn't--[p] THE WITNESS: For their purposes. [p] THE COURT: When you bet your life on it, you had better be certain. Now, there's no--as I read Frye and Kelley, there's no requirement that the Court make a finding that the scientific community certainly, without exception, absolutely, on the basis of their life, determines it to be reliable. [p] THE WITNESS: Um-hum, yeah, I read it the same way. [p] THE COURT: But when I read this chart, and there isn't one molecular biologist who has declared it to be not reliable--[p] THE WITNESS: Right, among the small group who thinks that they know something about it. THE COURT: Well, the small group happens to be molecular biologists that you're hopefully giving some credibility to in your overall survey. [p] THE WITNESS: Yeah, one has to be concerned that those in this group who knew something about the two companies differs in some way from the larger group, but that's--[p] THE COURT: Right. [p] THE WITNESS: Because it may be that the only people who know something about Lifecodes and Cellmark are people who have been involved in the companies, or have a different sort of experience than biologists in general, so--[p] THE COURT: That's an assumption on your part. [p] THE WITNESS: It's a possibility to keep in mind. [p] THE COURT: So, to get right down to my question again, does this draft of yours, being the face sheet of Defense 83, entitled, 'Reliability of Lifecodes and Cellmark Tests,' does it have any significance to you as it relates to the reliability of Lifecodes' and Cellmark's tests? [p] THE WITNESS: Certainly. I mean, it suggests that, among that group who claims to know about those tests, they all said it was reliable--at least in some version of reliability. [p] THE COURT: All right. Thank you. That was my question. That's all the questions I have. I have no other questions." 25 Dr. Mueller received his Ph.D. in ecology from University of California Davis. His specialty is the fruit fly. He testifies frequently for the defense in DNA cases including People v. Barney, supra, 8 Cal.App.4th 798, 10 Cal.Rptr.2d 731; People v. Axell, supra, 235 Cal.App.3d 836, 1 Cal.Rptr.2d 411; State v. Bible, supra, 175 Ariz. 549, 858 P.2d 1152. 26 This results when an individual inherits the same allele at a particular locus from his/her mother and father. If a person inherits different alleles from his/her mother and father, the person is said to be heterozygous. 27 Mr. McNeil is a registered medical technologist and is certified by the American Society of Clinical Pathologists. 28 Dr. Conneally received his bachelors from University College in Dublin and his masters and Ph.D. in statistics and human genetics from the University of Wisconsin. He is a charter member of the Human Genome Organization which maps human DNA and defines probes such as the ones used in RFLP testing. He was the director of the American Society of Human Geneticists for five years and is chairman of the genetics task force of the Muscular Dystrophy Association. 29 He received his Ph.D. in population genetics from Brown University and is a geneticist at the University of Florida in Miami. 30 To illustrate the "ceiling principle," let us say that the genetic markers studied include eye color and hair color and texture. Let us further say that the sample DNA pattern showed blue eyes and black, curly hair. Using the ceiling principle, one would chose from the three major data bases--Caucasian, Black, and Hispanic--the one with the highest frequency for each particular genetic marker. For eye color, the frequency of blue eyes would presumably be highest in the Caucasian data base; for black, curly hair, the most common frequency would most likely be found in the Black data base. If fewer than 10 percent of those in the data base had that banding pattern, a 10 percent "ceiling" would be used instead. At that point, one would multiply the frequency of blue eyes (from the Caucasian data base) by the frequency of black, curly hair (from the Black data base) to determine the likelihood, using the "ceiling principle," of finding someone having both blue eyes and black, curly hair. (This is known as the "product rule.") NOTE: This is only an analogy to make the "ceiling principle" more understandable--it is the "coding" parts of DNA that determine eye color and hair color and texture; RFLP analysis generally studies "noncoding" parts, or "junk DNA." 31 He has a bachelor's degree in psychology from Yale, a Ph.D. in psychology and math from Indiana University, and a J.D. from Emory. 32 He received his masters and Ph.D. in biophysics from the University of Pittsburgh. He teaches at Clemson University. 33 Dr. Ribeiro has a doctorate in physics and has been working at Brookhaven National Laboratory in the field of nuclear biophysics, specifically gel electrophoresis of nucleic acids. 34 This may no longer be Dr. Lander's position. He recently co-authored an article with Dr. Bruce Budowle in 371 Nature 735-738 entitled DNA fingerprinting dispute laid to rest, (October 27, 1994). The article is subtitled, "Two principals in the once-raging debate over forensic DNA typing conclude that the scientific issues have all been resolved." 35 Dr. Bakken received her Ph.D. in developmental genetics from the University of Iowa and is a professor at the University of Washington. 36 In view of the court's affirmation that it intended to read the entire transcript, we find insupportable defendant's speculation that the court, at the May 31, 1991, hearing "may not yet have read the preliminary hearing transcript, if it ever read that transcript." 37 Dr. Hagerman received his bachelor of science degree from the University of Oregon and his Ph.D. in biochemistry from Stanford. He has been on the faculty of the University of Colorado Health Sciences Center since 1980. 38 The court noted that "the scientific community generally is always growing and changing and learning new things." 39 The Attorney General asked this court to judicially notice the NRC report, and approximately 30 other scientific articles on DNA typing. Defendant also submitted the NRC report and various scientific articles to the court as exhibits and asked that they be filed. We took both requests under submission. We now grant both motions. 40 Barney was a consolidated appeal of two cases, People v. Barney and People v. Howard. Both involved the issue of whether DNA typing evidence met the standard for admissibility set forth in Kelly/ Frye. The laboratory performing the RFLP analysis in Barney was Cellmark; the FBI performed the RFLP analysis in Howard. 41 This statement could not be clearer. At least according to the NRC committee authors, the procedures used to declare a "match" have not lost the general acceptance of the scientific community in the post-Axell era. 42 We agree with this statement and with the holding in Axell that "since a match between two DNA samples means little without data on probability, the calculation of statistical probability is an integral part of the process and the underlying method of arriving at that calculation must pass muster under Kelly/ Frye." (People v. Axell, supra, 235 Cal.App.3d at pp. 866-867, 1 Cal.Rptr.2d 411; accord People v. Barney, supra, 8 Cal.App.4th at p. 817, 10 Cal.Rptr.2d 731; State v. Bible, supra, 175 Ariz. at p. 589, 858 P.2d at p. 1190.) 43 Under this method, scientists multiply together the frequencies with which each of the bands representative of a DNA fragment and visualized on the autoradiograms appears in the relevant population data base. 44 "When there is no correlation between the two parental alleles, the locus is said to be in Hardy-Weinberg equilibrium." (NRC Report at p. 78.) 45 "[T]he absence of [a correlation between genotypes at different loci] is called linkage equilibrium." (NRC Report at. p. 78.) 46 For an explication of the "ceiling principle," see footnote 30. 47 Barney has been followed in out-of-state cases, including State v. Bible, supra, 175 Ariz. at 579, 858 P.2d at 1182; U.S. v. Porter (D.C.App.1992) 618 A.2d 629, disapproved for retroactively considering the NRC Report in U.S. v. Bonds (6th Cir.1993) 12 F.3d 540, 553; and People v. Watson (1 Dist.1994) 257 Ill.App.3d 915, 196 Ill.Dec. 89, 629 N.E.2d 634, also disapproved in U.S. v. Bonds, supra. The only California case following it, People v. Wallace (1993) 14 Cal.App.4th 651, 17 Cal.Rptr.2d 721, was authored by the same justice who wrote Barney. On the other hand, two federal cases (U.S. v. Bonds, supra, 12 F.3d at p. 564 and U.S. v. Chischilly (9th Cir.1994) 30 F.3d 1144, 1154) have criticized Barney. 48 Lander now claims that the scientific issues have all been resolved. See footnote 34, ante. 49 We express no opinion on the scientific acceptance of other processes of DNA typing, such as polymerase chain reaction (PCR), not involved in this case. 50 We do not know how many scientific articles were before the court in Barney. In this case, however, defendant and the attorney general have supplied this court with approximately 30 scientific articles each, which have been admitted as exhibits. This court has read and considered the articles submitted. In addition, dozens of scientific articles were admitted below, either in connection with the Kelly/ Frye hearing, the discovery motions, or trial. Finally, this court has independently examined further scientific journals. 51 The government called six expert witnesses and the defense five. In addition, the court called Dr. Eric Lander (see fn. 34) as the court's witness. 52 In this trial, Dr. Conneally described Lewontin and Hartl as "fruit fly geneticists" who "are not necessarily as qualified to talk about human populations...." 53 We reiterate that this is not the unanimous view of the scientific community, but unanimity is not required. (See People v. Leahy, supra, 8 Cal.4th at p. 611, 34 Cal.Rptr.2d 663, 882 P.2d 321; People v. Guerra, supra, 37 Cal.3d at p. 418, 208 Cal.Rptr. 162, 690 P.2d 635.) Critics of the use of DNA to inculpate (but not to exculpate) a defendant include many of the defense experts who testified here, Lewontin and Hartl, Dr. Joel M. Cohen, and Dr. Seymour Geisser. 54 "The second step of DNA analysis is to compare the DNA patterns produced by the processing step in order to determine whether the suspect's DNA pattern matches the DNA pattern of bodily material found at the crime scene. [p] First, the patterns are visually evaluated (i.e., 'eyeballed') to determine whether there is a likely match. Most exclusions will be obvious, since the patterns will be noticeably different. If there is not an obvious exclusion, the bands in the patterns are subjected to computer-assisted analysis to determine the length of the represented DNA fragments as measured in base-pair units. The measurements are taken by comparing the bands for the sample fragments with the bands for the size-marker fragments of known base-pair lengths. [p] Because of inherent limitations in the DNA processing system, it is not possible to measure the sample fragments to the precise number of base pairs. Thus, a margin of error is built into the matching system. Fragments being compared need not have been assigned precisely the same base-pair length for a match to be declared. For example, the FBI laboratory will declare a match within a 'match window' if two fragments differ in base-pair length by less than 2.5 percent. [p] It is not particularly uncommon for two people to share the same lengths for two fragments, but it is very unusual to share the same lengths for all eight fragments. Thus, if all of the suspect's fragment lengths are the same as the crime scene fragment lengths within the margin of error--i.e., if the band patterns produced by the processing step are identical--a match is declared." (People v. Barney, supra, 8 Cal.App.4th at pp. 808-809, 10 Cal.Rptr.2d 731.) 55 But see the Fifth District case, People v. Pizarro (1992) 10 Cal.App.4th 57, 12 Cal.Rptr.2d 436 (remand for further hearing on whether FBI's testing procedures and population data base were generally accepted in the scientific community; court noted that Cellmark's procedure had been found to be generally accepted but there was no evidence presented that FBI's methodology was essentially the same as Cellmark's). ** See footnote *, ante.